DE102019108856A1 - Cross-database index on a distributed database system - Google Patents

Abstract

The invention relates to a computer-implemented method for implementing a cross-database index (112, 222) on a distributed database system (170) which comprises a plurality of independent individual databases (104, 204). The individual databases are communicatively connected to one another via a network (180). The method comprises applying a data processing function of a source database (104) to an additional data record (108), the data processing function comprising at least indexing the additional data record for storage in the source database. Furthermore, a supplementary data record (130) resulting from the use of the data processing function is created, which includes the additions made in the index of the source database and serves to supplement at least one further index (212) of at least one further individual database of the distributed database system as a receive database (204). The supplementary data set created is sent via the network to the second individual database for integration into the further index of the receiving database (204).

Description

The invention relates to a method, a source computer system and a distributed database system for implementing a cross-database index on a distributed database system.

In the course of increasing digitization, automation and networking in all areas of life and work, complex distributed systems are emerging, the technical interaction of which raises new technical challenges. For example, corresponding distributed systems generally do not include a shared memory, which makes it difficult to carry out data processing functions.

The invention is based on the object of creating an improved method for implementing an index on a distributed database system.

The object on which the invention is based is achieved in each case with the features of the independent patent claims. Embodiments of the invention are specified in the dependent claims.

• • Empfangen eines zusätzlichen Datensatzes durch eine erste Einzeldatenbank als Quelldatenbank zum Ergänzen von Datensätze der Quelldatenbank,
• • Speichern des zusätzlichen Datensatzes, welcher ein oder mehrere zusätzliche Feldwerte umfasst, durch ein erstes Multi-Modell-Datenbankmanagementsystem der Quelldatenbank in einem dokumentenorientierten ersten Datenmodell der Quelldatenbank,
• • Anwenden einer Datenverarbeitungsfunktion der Quelldatenbank auf den zusätzlichen Datensatz, wobei die Datenverarbeitungsfunktion zumindest eine Indexierung des zusätzlichen Datensatzes zum Speichern in einem zweiten Datenmodell der Quelldatenbank umfasst, wobei die Indexierung umfasst:
  • ◯ Erzeugen von ein oder mehreren zusätzlichen Token aus den zusätzlichen Feldwerten durch die Quelldatenbank,
  • ◯ Ergänzen des Index der Quelldatenbank durch das erste Multi-Modell-Datenbankmanagementsystem unter Verwendung der zusätzlichen Token und eines Zeigers auf den zusätzlichen in dem dokumentenorientierten Datenmodell der Quelldatenbank gespeicherten Datensatz,
• • Erstellen eines aus dem Anwenden der Datenverarbeitungsfunktion resultierenden Ergänzungsdatensatzes, welcher die in dem Index der Quelldatenbank vorgenommenen Ergänzungen umfasst, zum Ergänzen zumindest eines zweiten Index zumindest einer zweiten Einzeldatenbank des verteilten Datenbanksystems als Empfangsdatenbank,
• • Senden des Ergänzungsdatensatzes über das Netzwerk an die zweite Einzeldatenbank zur Integration in den zweiten Index der Empfangsdatenbank.

Embodiments include a computer-implemented method for implementing a cross-database index on a distributed database system which comprises a plurality of independent individual databases, the individual databases being communicatively connected to one another via a network,
the individual databases each being managed by a multi-model database management system, the individual databases each comprising a plurality of database-specific data records which are stored in a document-oriented first data model of the respective individual database, the stored data records each comprising one or more field values, the individual Field values of the saved data records are each saved in a field,
wherein the individual databases further each comprise a searchable first index which is stored in a second data model of the respective individual database, the index of the respective individual database comprising a plurality of tokens generated from the field values of the data records stored in the document-oriented data model of the corresponding individual database, the Token in the index is linked with one or more pointers to one or more of the data records stored in the document-oriented data model of the corresponding individual database, from whose field values the corresponding token was generated,
the method comprising:

• • Receipt of an additional data record by a first individual database as a source database to supplement data records from the source database,
• • Saving the additional data record, which includes one or more additional field values, by a first multi-model database management system of the source database in a document-oriented first data model of the source database,
• Applying a data processing function of the source database to the additional data record, the data processing function comprising at least one indexing of the additional data record for storage in a second data model of the source database, the indexing comprising:
  • ◯ Generation of one or more additional tokens from the additional field values by the source database,
  • ◯ Supplementing the index of the source database by the first multi-model database management system using the additional tokens and a pointer to the additional data set stored in the document-oriented data model of the source database,
• • Creation of a supplementary data record resulting from the use of the data processing function, which comprises the additions made in the index of the source database, for supplementing at least one second index of at least one second individual database of the distributed database system as a receive database,
• • Sending the supplementary data set via the network to the second individual database for integration into the second index of the receiving database.

Embodiments can have the advantage that by creating and sending the supplementary data record, the index of the receive database can be adapted to the changes in the index of the source database. Even though it is a distributed database system with a plurality of independent individual databases, a cross-database index can be implemented in this way. If all individual databases are configured to function as a source database when an additional data record is received, a uniform cross-database index of the distributed database system can be provided. If all the individual databases are configured to receive appropriate supplementary data records as receive databases, an identical cross-database index of the distributed database system can be used on all individual databases to be provided. This allows data processing functions to be carried out locally, taking into account all the data made available in the distributed database system. At the same time, however, not all individual databases have access to all the data available in the distributed database system. Rather, access is only made to the cross-database index. Due to the tokenization, the index does not include the full context information that the underlying data records include. However, this full context information is not lost, but is available to the individual databases in each case. Due to the reduced and compressed information content of the index, on the one hand, the amounts of data that have to be exchanged between the individual databases to implement the cross-database index are reduced and, at the same time, the data in the data records are protected. This protection includes that only the individual, independent individual databases have access to the data records stored in their respective first data model. Only this respective individual database therefore has access to the full information content of the corresponding data records. However, because of the pointers, the other individual databases have knowledge of where data records are stored from which the tokens de index were generated. This means you can request a copy of the relevant data records if necessary. The decision about releasing a data record, ie providing a copy of the same, however, remains in each case with the corresponding individual database in whose first data model the data record is stored. The power of disposal therefore lies with the individual database. Therefore, the individual databases are also independent in the sense that each of the individual databases in its first data model each comprises an independent plurality of data records, which may differ significantly or even completely from data records that are stored in the first data models of other individual databases. In particular, the restricted access protects against unauthorized access to the data by third parties.

Embodiments can have the advantage that the data processing functions are each carried out individually in a single database. For example, each of the individual databases independently indexes data records that it receives. As a result, the data records can be processed in parallel. Thus, the processing of the data can be parallelized and thus accelerated.

Embodiments are thus based on a distributed system or database system, which comprises a plurality of nodes or computers, which each provide an individual database and can communicate with one another via communication interfaces. According to embodiments, software, i. E. the data processing functions, which can provide a plurality of functionalities such as algorithms, formulas, models.

According to embodiments, the data processing function can exhaust itself from indexing received data records and creating or managing an index based thereon. According to further embodiments, however, the data processing function can also include further functionalities, the application of which is found in the resulting structure of the index. The data provided by the index can thus be optimized for a variety of purposes.

The software can, for example, be software which is installed on the individual nodes of the system, i.e. the individual databases, independently of the other nodes of the distributed system. The distributed system in this case refers to the fact that each of the nodes of the distributed system includes a complete copy of the index. The resulting index indexes all individual databases of the distributed system, i.e. the index can be a combined overall index which includes all sub-indexes of the individual accounts of the distributed system.

A distributed application describes a complex application program that runs on a distributed system, i.e. on several computers, and exchanges information between them. For example, the corresponding application is distributed through horizontal cuts in the software layer model, so that the task of the overall software can be divided into individual software components in the form of the individual cuts. In this case, all components of the application that communicate with one another for this purpose must cooperate in order to fulfill the overall task. The distributed application is mostly transparent to a client, i.e. the distributed application appears as a single application.

According to embodiments, each individual database carries out a data processing function. This can be the same data processing function in all individual databases, which is therefore parallelized, or it can be
act single database-individual data processing functions, the execution of which is complementary. For example, the creation of an index of different individual databases can change this differentiate the granularity of the tokens added to the index. In the case of sensor measured values, for example, the exact measured values can be added to the index, a rounded measured value or assignments to a value interval can be added. In the case of values from different sensors, there may be different requirements for them, which can be implemented by means of individual database-specific data processing functions.

A distributed system is understood to mean a plurality of interacting processes or a plurality of interacting processors executing processes which do not have a common memory, but rather communicate with one another via messages. A distributed system thus enables independent computers or individual databases to be merged, each of which is presented as a single system for the respective user. In the present case, the respective user sees, for example, a cross-database index. Only when data records are to be accessed directly, which are stored in a data model of another individual database, can it become apparent to the user that it is not a single system but rather a distributed system.

Embodiments can have the advantage that the functionality of the complete index can be used on each of the nodes or on each of the individual databases without the corresponding nodes including all data records of the entire distributed database system. As a result, an overhead for the operation and storage of the data records of the remaining nodes or individual databases can be avoided on the respective nodes.

Embodiments can have the advantage that a complete central storage of all data or a merging of all databases is not necessary at any location so that the entirety of the information of all indexes of all databases can be accessed and evaluated. This prevents a central instance from collecting all the data that could access all of the data. In particular, associated security risks can be avoided or minimized.

Embodiments can have the advantage that data which are recorded by a node in the distributed network are linked directly to an owner or an identity. For example, the received data records are each assigned directly to the received individual database or to a user of the corresponding individual database as the owner or owner.

Embodiments can have the advantage that they enable functionalities of a software component to be operated in a targeted manner on individual network nodes in a distributed system. In this way, the memory and computing requirements on the relevant node can be minimized while the desired functionality is operated. This also enables relatively underperforming, e.g. in relation to other nodes of the distributed network, to execute functionalities locally on site.

Embodiments may have the advantage of providing an implementation of an edge computing system.

Edge computing describes decentralized data processing at the edge of a network, the so-called edge. Edge computing is in contrast to cloud computing, in which computer applications, data and services are centralized, i.e. from central network node nodes, a plurality of decentralized network nodes are provided for use.

In contrast, edge computing is geared towards processing computer applications, data and services locally or at least in the vicinity if possible. This can have the advantage that data streams can be limited to local network areas in a way that conserves resources. In particular, resources can be used that are not permanently connected to a network, in particular mobile computer systems, e.g. Controllers, notebooks, smartphones, tablet computers and sensors. Different processes and structures can be used in edge computing, such as sensor networks, mobile data acquisition, mobile signature analysis and / or peer-to-peer and ad-hoc networking. Edge computing can have the advantage that it can be used as an architectural concept for the Internet of Things (loT), which allows a link between clearly identifiable physical objects, i.e. "Things", with a virtual representation or identity in a network structure, e.g. an "Internet" -like structure.

Services that are provided in the course of edge computing can have the advantage that the data volume to be transmitted and thus the data exchange as well as the transmission paths can be significantly reduced, which means that transmission costs and waiting times can be reduced. As a result, the overall service quality can be improved. With edge computing, central data centers are either not or to a large extent in small number is necessary, whereby bottleneck effects at such data centers for data transfer can be avoided as well as the associated potential sources of error. By processing encrypted files closer to the respective data sources and / or encryption instances, the security of the data within the distributed system can also be increased. In particular, this enables potential security threats such as viruses, falsified data and / or hacker attacks to be recognized early and effectively averted. In addition, the ability to virtualize extends the scalability within the system, ie the number of edge devices in the network can be easily increased. Finally, when using edge computing, real-time requirements in the Internet of Things can be better supported as a result of faster data processing than is generally possible in the course of cloud computing.

In everyday life, users increasingly surround themselves with electronic devices that have computer functionality and connectivity. One trend is the increasing spread of wearables, i.e. Computer systems which are arranged on the body of the user during use. Another trend is the replacement of everyday objects with “intelligent objects” (also known as “smart devices”), which are equipped with information technology and are configured to process information themselves. These intelligent objects are equipped with data processing hardware, such as a built-in microcontroller, communication interfaces and / or sensors, so that they can record, store and / or exchange data with one another. A catchphrase for this development is, for example, the so-called Internet of Things (also abbreviated to loT for “internet of things”).

For example, a plurality of data acquisition devices are provided, which are arranged in the distributed database system, acquire or generate data locally and transmit this data or supplementary data records based thereon to one or more other nodes via a communication interface over the network.

By reducing the software to its core functionalities, memory requirements and computing capacity can be saved so that the corresponding software can be used effectively and efficiently in loT sensors, for example. Linking the data with an owner or with an identity increases the level of security of the data, since subsequent manipulation of the data by unauthorized persons can be excluded.

By using specific functionalities of a software component, functionalities can be executed locally on resource-restricted nodes in a distributed system, for which the entire software would otherwise be necessary.

Embodiments can have the advantage that they can provide an advantageous distributed database system for the Internet of Things.

• • Abgleichen der zusätzlichen Token mit dem Index der Quelldatenbank,
• • falls eines der zusätzlichen Token nicht von dem Index der Quelldatenbank umfasst ist, Ergänzen des entsprechenden zusätzlichen Tokens in dem Index der Quelldatenbank und Verknüpfen des entsprechenden zusätzlichen Tokens in dem Index mit dem Zeiger auf den zusätzlichen in dem dokumentenorientierten Datenmodell der Quelldatenbank gespeicherten Datensatz,
• • falls eines der zusätzlichen Token von dem Index der Quelldatenbank umfasst ist, Verknüpfen des entsprechenden zusätzlichen Tokens in dem Index der Quelldatenbank mit dem Zeiger auf den zusätzlichen in dem dokumentenorientierten Datenmodell der Quelldatenbank gespeicherten Datensatz.

According to embodiments, supplementing the index includes:

• • Matching the additional tokens with the index of the source database,
• • if one of the additional tokens is not included in the index of the source database, adding the corresponding additional token in the index of the source database and linking the corresponding additional token in the index with the pointer to the additional data set stored in the document-oriented data model of the source database,
• If one of the additional tokens is included in the index of the source database, linking the corresponding additional token in the index of the source database with the pointer to the additional data record stored in the document-oriented data model of the source database.

Embodiments can have the advantage that data from additional data records can be efficiently inserted into the existing individual database and in particular into the index. The tokens generated using the additional data record are compared with the index. All tokens that the index does not (yet) include are added to the index as additional tokens. Furthermore, the additional tokens are each linked to the pointer to the additional data record. Thus, even when the corresponding index structures are transferred to supplement another index of another individual database, the pointers can be used to determine which individual database comprises the underlying data or data records.

For those tokens of the additional data record which the index already includes, the pointer to the additional data record is added to the index.

Embodiments can have the advantage that it can always be ensured that the index contains all of the data records in the database tokens included or at least taken into account.

According to embodiments, the supplementary data set is sent from the source database via the network to a predefined first group of one or more individual databases of the plurality of individual databases as receive databases which the second individual database comprises. Embodiments can have the advantage that it can be ensured for a predefined first group that its members each have the complete cross-database index. This can be advantageous, for example, in an asymmetrical system in which there is no need to provide a complete index for one or more of the individual databases. This can be the case, for example, if no or only seldom local searches are carried out on these databases. In this case, the corresponding individual database serves primarily as a data store and for parallelizing the creation of the index.

According to embodiments, the supplementary data set is sent from the source database via the network to all further individual databases comprised by the plurality of individual databases as receive databases. Embodiments can have the advantage that a cross-database index can be implemented on each of the individual databases.

According to embodiments, a predefined second group of a plurality of individual databases of the plurality of individual databases is configured to function as a source database in each case when additional data records are received. According to embodiments, all of the individual databases of the plurality of individual databases are configured to act as a source database when additional data records are received. Depending on how many of the individual databases are also configured as source databases, different degrees of standardization of the individual indexes of the individual databases can be achieved.

According to embodiments, a sensor is assigned to the source database and the additional data record comprises sensor data recorded by the sensor. Embodiments can have the advantage that the sensor data are indexed directly after their acquisition. According to embodiments, each individual database of the plurality of individual databases is assigned one or more sensors, from which the respective individual database receives recorded sensor data in the form of additional data records.

According to embodiments, the data processing function further comprises normalizing the additional tokens in the course of adding to the index of the source database. According to embodiments, the normalization fulfills the fifth and / or sixth normal form. Embodiments can have the advantage that redundancies can be avoided.

According to embodiments, the tokens can be stored in the form of relations or equivalent structures. In the sense of relational database theory, a relation is a set of tuples. A tuple is a set of attribute values. An attribute describes a data type or a property assigned to one or more data. The number of attributes determines the degree, the number of tuples the cardinality of a relation.

Normalization, in particular normalization of a relational data model, is understood to mean a division of attributes into a plurality of relations according to a normalization rule, so that redundancies are reduced or minimized. A relational data model can be implemented, for example, in table-like data structures in which the relations are implemented in the form of tables, the attributes in the form of table columns and the tuples in the form of table rows.

Data redundancies run the risk of inconsistencies and anomalies occurring when data is changed that is multiple times. Furthermore, redundancies unnecessarily increase the storage space requirement. Such redundancies can be reduced or minimized by normalization. A relational data model can be brought into a normal form, for example, in that the relations of the data schema are progressively broken down into simpler relations on the basis of the functional dependencies applicable to the corresponding normal form.

For example, the following normal forms can be distinguished: 1. normal form (1NF), 2. normal form (2NF), 3 . Normal form (3NF), Boyce-Codd normal form (BCNF), 4th normal form (4NF), 5th normal form (5NF), 6th normal form (6NF).
The normalization criteria increase from normal form to normal form and each include the normalization criteria of the previous normal forms, ie 1 NF ⊆ 2NF ⊆ 3NF ⊆ BCNF ⊆ 4NF ⊆ 5NF ⊆ 6NF.

A relation is in the first normal form if each attribute of the relation has an atomic range of values and the relation is free of repeating groups. The term atomic is used here to exclude composite, quantity-valued or nested value ranges for the attributes, i.e., relation-valued attribute value ranges, Understood. Freedom from repeating groups requires that attributes that contain the same or similar information are transferred to different relations.

A relation is in the second normal form if it meets the requirements of the first normal form and no non-primary attribute is functionally dependent on a real subset of a key candidate. A non-primary attribute is an attribute that is not part of a key candidate. This means that each non-primary attribute is dependent on all whole keys and not just on part of a key. Relations in the first normal form, the key candidates of which are not composed but consist of a single attribute, therefore automatically satisfy the second normal form. A key candidate is understood here to be a minimal set of attributes that uniquely identify the tuples of a relation.

A relation is in the third normal form if it fulfills the requirements of the second normal form and no non-key attribute depends transitively on a key candidate. An attribute is transitively dependent on a key candidate if the corresponding attribute is dependent on the corresponding key candidate via a further attribute.

A relation is in Boyce-Codd normal form if it meets the requirements of the third normal form and every determinant is a super key. A determinant is understood here to be a set of attributes on which other attributes are functionally dependent. A determinant thus describes the dependency between attributes of a relation and defines which attribute sets determine the value of the other attributes. A super key is a set of attributes in a relation which uniquely identify the tuples in this relation. The attributes of this set therefore always include different values in the case of tuples selected in pairs. The key candidate is therefore a minimal subset of the attributes of a super key, which enables the tuple to be identified.

A relation is in the fourth normal form if it meets the requirements of the Boyce-Codd normal form and does not contain any nontrivial multivalued dependencies.

A relation is in the fifth normal form if it meets the requirements of the fourth normal form and does not include any multivalued dependencies that are mutually dependent. The fifth normal form is therefore present if every non-trivial compound dependency is implied by the key candidates. A compound dependency is implied by the key candidates of the output relation if each relation of the set of relations is a superkey of the output relation.

A relation is in the sixth normal form if it fulfills the requirements of the fifth normal form and does not contain any nontrivial join dependencies. A relation is sufficient for a join dependency of a plurality of relations if the relation as the starting relation can be broken down into the corresponding set of relations without loss. The link dependency is trivial if one of the relations of the set of relations has all the attributes of the output relation.

According to embodiments, the data processing function further comprises an assignment of rights to manage and / or process the data of the additional data record. In particular, an owner can be assigned directly to the additional data record in this context. This owner can, for example, be given certain access rights at the same time, which enables him to manage and / or process the data. According to embodiments, the first multi-model database management system assigns an owner right with regard to the additional data record and / or the supplementary data record to a first entity assigned to the source database. The corresponding entity is thus defined as the owner or owner of the additional data record and / or the supplementary data record.

According to embodiments, the data are linked directly to an owner or assigned to an owner by the node that acquires or generates the corresponding data.

According to embodiments, the collected or generated data is used on the resource-constrained nodes of the distributed system, e.g. loT sensors, authorization chains formed with the collected data. The authorization chains define access rights for a number of entities. According to embodiments, a central database with an authorization concept is not necessary on these systems, since the functionalities of the authorization chains are implemented directly on the individual nodes. As a result, data can be effectively and efficiently linked to an owner even on resource-constrained nodes.

Linking the data with an owner directly when data is generated on the individual resource-restricted nodes of the distributed system can have the advantage of increasing data security and ensuring transparency about the actual owners of the respective data.

According to embodiments, the first entity is a first user of the source database or a first computer system assigned to the source database. If it is an assigned computer system, e.g. comprising the source database, this can be advantageous for applications in the field of the Internet of Things, as each data record recorded by a corresponding sensor system is automatically assigned a unique owner during the course of entering it into the corresponding individual database.

• • ein Lesezugriffs-Berechtigungsnachweis, welcher einen lesenden Zugriff auf den Inhalt eines Datensatzes ermöglicht;
• • ein Schreibzugriffs- Berechtigungsnachweis, welcher modifizierenden Zugriff auf den Inhalt eines Datensatzes ermöglicht;
• • ein Indexzugriffs- Berechtigungsnachweis, welcher Kenntnis der Existenz des Datensatzes in der den Datensatz beinhaltenden Datenbank und einen lesenden Zugriff auf Metadaten des Datensatzes ermöglicht.

According to embodiments, access authorization certificates are stored in the source database as part of the additional data record and / or the supplementary data record, the access authorization certificates comprising one or more of the following access authorization certificates:

• • a read access authorization, which enables read access to the content of a data record;
• • a write access credential which enables modifying access to the content of a data record;
• • An index access authorization certificate, which enables knowledge of the existence of the data record in the database containing the data record and read access to metadata of the data record.

Access credentials can be designed as access certificates. According to embodiments, the access certificates are pure numeric values, not complex x509 certificates. Metadata regarding the validity of the certificates and regarding other aspects can be stored separately from the actual access certificate in an additional ID database of an access management system of the corresponding individual database. Each data record preferably contains in its corresponding fields all access certificates of the user creating this data record or of the entity to which this data record is inertially assigned. For example, if a data record is DS from user U1 created and the user U1 has exactly according to the content of the ID database 3 Types of certificates (a read access certificate "U1.Z-Zert [R]", a write-access certificate "U1.Z-Zert [W]" and an index access certificate "U1.Z-Zert [S]"), so are copies of just that 3 Access certificates created in the course of storing the data record DS from the ID database and stored in the corresponding fields of the data record DS. With the user U1 it can also be the corresponding individual database itself or an identity assigned to this individual database. For example, a secret key of an asymmetrical cryptographic key pair assigned to the corresponding individual database can serve as proof of authentication or proof of identity. Alternatively, the identity can also be based on technical characteristics, such as certain voltage or current patterns of the corresponding individual database or of a computer system comprising the corresponding individual database, comparable to biometric features in the case of living people.

According to embodiments, the individual database and / or components of the same have means for protection against unauthorized manipulation. The means for protection against unauthorized manipulation ensure the trustworthiness of the individual database or its control elements, such as processors, that is to say its function as a “trust anchor”, through technical measures. For example, the individual database comprises components or is implemented on components which are provided by a trustworthy institution, e.g. configured by a trust center (trust service provider) and provided with the required cryptographic key material. The means for protection against unauthorized manipulation can ensure that security-relevant functionalities of the individual database or of the computer system comprising this individual database are not modified in an unauthorized manner.

According to one embodiment of the invention, the means for protecting against unauthorized manipulation are designed as a so-called tamper proof module or trusted platform module (TPM), which is also referred to as a tamper resistant module (TRM). For example, at least parts of the individual database or of the computer system comprising the individual database are signed, such as program components and / or hardware components that can carry a digital signature. In particular, an operating system, multi-model database management system, a configuration file and / or a mass storage medium can be digitally signed. Before using the individual database, the TRM then checks whether the signature or the signatures are valid. The test can also be based on the technical characteristics mentioned above. If one of the signatures is not valid, the TRM blocks the use of the individual database. A TPM includes, for example, microcontrollers according to the TCG specification as in ISO / IEC 11889 that adds basic security functions to a computer or similar device.

According to one embodiment of the invention, the means for protection against unauthorized manipulation include mechanical means which, for example, are intended to prevent components and / or the entire computer system from being opened or which, when an attempt is made to intervene, lead to the corresponding components become unusable, for example if a data loss and / or a data block occurs. For this purpose, for example, safety-critical parts can be cast in epoxy resin, an attempt to remove a relevant component from the epoxy resin leads to the inevitable destruction of this component. In particular, the means for protecting against unauthorized manipulation can be designed as a so-called hardware security module (HSM).

If now the user U1 the first individual database to another user U2 grants the same individual database or a further individual database or a further individual database per se read rights with regard to the data record DS, this means that a link or assignment of this read access certificate "Z.Zert_U1 [R]" of User U1 and the user's user certificate U2 is stored, thereby forming an authorization chain object. If the user U2 now wants to access the data record DS at a later point in time, sends the individual database containing the data record DS in response to the access request of the user U2 automatically sends an authorization request to the ID database of the access management system. This request contains the access rights of the creator stored in the data record DS U1 . The ID database then checks whether the user U2 assigned user certificate in the ID database with one or more of the author's access rights stored in the data record DS U1 linked is saved. Only if this is the case, and if the user U2 is also the owner of the source database, he is allowed to access the data record.

According to embodiments, the at least one access certificate, which is preferably stored as part of the data record created, comprises several access certificates for other types of access. The multiple access certificates include, for example, a write access certificate Z.Zert_U1 [W] of the creating user and / or a read access certificate Z.Zert_U1 [R] of the creating user and / or an index access certificate Z.Zert_U1 [S] of the creating user.

According to embodiments, the access management system is a “standard” access management system that is usually installed on a standard computer system. A "standard" access management system can e.g. MySQL, PostGreSQL, Oracle, SAP Hana, etc. However, it is also possible that other systems, which are designed for the structured storage and retrieval of data, are used as the access management system, e.g. Microcontroller, in whose memory the data records, certificates and authorization chain objects as well as the program logic described here can be stored.

According to embodiments, a user certificate is, for example, a certificate that is issued specifically for a particular person by a certification authority (CA). The user certificate is arranged in a testable manner in a certificate chain issued by a certification authority, so it can be tested for example up to the root certificate of the certification authority. This can be advantageous because certification bodies are already widely accepted as independent guarantors of trust and are already used by many existing technical systems to check the authenticity of certain users and user actions.

According to embodiments, a plurality of user certificates and / or a plurality of access certificates are stored in an ID database of the access management system. The ID databases also contain access authorization chain objects.

An authorization chain object, more precisely an access authorization chain object, is a data object which contains one of the access certificates and one or more of the user certificates (and thereby assigns them to one another). The sequence of the user certificates in the access authorization chain object reflects the sequence of users who have issued this access certificate for other users whose user certificate is contained in the access authorization chain object. An authorization hierarchy generated in this way comprises one of the access authorization chain objects.

According to embodiments, each of the individual databases comprises a corresponding ID database which manages the access authorizations to the data records and / or the index of the corresponding individual database.

According to embodiments, the individual databases or their data models are free of access authorization chain objects and only contain the access certificates for each data record that grant the user who has created this data record or to whom this data record is assigned access to this data record.

The linking of these access rights of the record producer with one or more other users in order to grant this access to the record is not stored in the individual databases or their data models, but in the ID database. Conversely, the ID database does not contain any reference to individual data records in the individual database or its data models. This can be beneficial since the size and complexity of the individual data records of the individual databases or their data models are limited and logically largely decoupled from the administration of the access rights. The size of the data records is also limited by the fact that the complete chain of authorization transfers is not saved as part of the data records. This can considerably reduce the storage space required by the database, particularly when there are a large number of small data records with an identical authorization structure.

According to embodiments, the ID database of the access management system includes a private signing key. The individual database contains a public signature verification key, which is designed to verify the signatures created with the signature key. When a first user requests access to a data record in the individual database, the access management system determines whether the user certificate in the ID database is assigned one or more access authorization certificates, a copy of which is stored as part of the data record. If this is the case, this proof of access authorization is signed with the signature key of the ID database and checked with the corresponding signature verification key of the individual database. The credentials are sent to the individual database in signed form. The individual database uses the signature verification key to check whether the signature of the proof of authorization is valid, the establishment of the database connection and data record access being permitted only if the signature is valid.

For example, a central ID database is provided for the entire distributed database system. This can be advantageous because it is not necessary because not every single database has to be equipped with program logic for the certificate chain check. In general, the described method or the described database structure can be used to achieve an extensive separation of data management (in individual databases) and the management of access rights (in the ID database). The ID database contains and / or is operatively linked to one or more program modules, for example for certificate chain checking (for example from user certificates to the root certificate from the certification authority issuing the user certificate), for storing changes to the assignments of Certificates and users in a log file as well as for the dynamic creation of signed credentials, taking into account a documented chronology of the transfer of rights. In contrast, according to embodiments, the individual individual databases only have means for checking whether corresponding authorizations exist for the individual database itself or for the data records contained therein or are operatively linked to them, these means possibly including means for signature checking.

According to one embodiment, each of the data records in the source database contains one or more access certificates of the user creating the data record or of an entity to which the corresponding data record is assigned. The access certificates in the ID database are each assigned one or more user certificates so that the chronological sequence of users who have granted themselves one or more of the access rights of the user creating the data record is represented in the form of a hierarchy. This assignment can e.g. by means of access authorization chain objects.

In the following, an “access management system” is understood to mean an electronic system for storing and retrieving data. For example, the access management system can be a “classic database management system” (DBMS) (MySQL, PostgreSQL, Oracle, Hana, etc.). But it is also possible that the data are stored in a microcontroller memory and managed by an application program or a chip-based program logic that works as an access management system and is not a classic DBMS. The data are preferably stored consistently and permanently in the access management system and are efficiently made available to various application programs and users in a form that meets their needs. An access management system can typically contain one or more databases and manage the data records contained therein.

In the following, an “ID database” is understood to mean a database which contains and manages user-related information such as user certificates and the rights assigned to these users in the form of further certificates (access certificates). In contrast to the individual databases, which can also be summarized under the term "user data databases" and which are primarily used to store user data, an ID database is primarily used to manage the owner and access rights assigned to users with regard to user data.

According to embodiments, at least one index access credential is stored as part of the additional data record and / or the supplementary data record for the first entity. According to embodiments, as part of the additional data record and / or the supplementary data record for the first entity, a Read access credentials, write access credentials and index access credentials are stored.

According to embodiments, an index access credential for the additional data record is also assigned to a second entity assigned to the receive database. This can be done, for example, by means of an access authorization chain object described above. According to embodiments, the second entity is a second user of the receive database or a second computer system assigned to the receive data.

According to embodiments, a prerequisite for sending the supplementary data record to the receiving database is a successful check that the second entity is assigned an index access authorization for the supplementary data record. It can thus be ensured that the receive database is actually authorized to access corresponding index information such as the supplementary data record.

According to embodiments, the data processing function further comprises a classification of the tokens generated from the additional data record.

• • falls eines oder mehrere erste zusätzliche Token von einer der Trigger-Definitionen als Trigger umfasst sind, Identifizieren des entsprechenden Tokens als Trigger durch das Lernmodul,
• • Verwenden der identifizierten Trigger zum Zuordnen von ein oder mehreren zweiten zusätzlichen Token zu ein oder mehreren Klassen der ersten Gruppe von Klassen durch das Lernmodul, falls die entsprechenden zweiten zusätzlichen Token von dem zusätzliche Datensatz in einer Kombination mit ein oder mehreren der identifizierten Trigger gemäß einer der Trigger-Definitionen umfasst sind, wobei die entsprechenden Trigger gemäß der entsprechenden Trigger-Definition eine entsprechende Klassenzuordnung triggern, wobei das Ergänzen des Index durch das erste Multi-Modell-Datenbankmanagementsystem unter Verwendung der Klassenzuordnungen der zusätzlichen Token erfolgt.

According to embodiments, the source database for the classification comprises a pre-trained learning module for machine learning, the pre-trained learning module comprising a plurality of predetermined trigger definitions which define triggers for assigning tokens to classes of a first group of classes,
wherein first tokens in the index of the source database, which are included in one of the trigger definitions of the source database as a trigger, are each assigned to the corresponding trigger definition, wherein second tokens in the index of the source database each have one or more classes of the first group of classes assigned,
wherein the classifying comprises:

• • if one or more first additional tokens are included in one of the trigger definitions as a trigger, the learning module identifies the corresponding token as a trigger,
• Use the identified triggers to assign one or more second additional tokens to one or more classes of the first group of classes by the learning module if the corresponding second additional tokens from the additional data set in a combination with one or more of the identified triggers according to a the trigger definitions are included, the corresponding triggers triggering a corresponding class assignment according to the corresponding trigger definition, the addition of the index by the first multi-model database management system using the class assignments of the additional tokens.

Embodiments can have the advantage that the learning module is a pre-trained learning module. The pre-trained learning module comprises a plurality of inertially provided or defined trigger definitions. The learning module is configured to classify all tokens included in the database or the index using these inertially defined trigger definitions. Embodiments can have the advantage that no chance enters the decision-making or classification process. Rather, the classification of tokens is based on the predetermined trigger definitions and can therefore be traced at any time. Even if the learning module advances, for example, on the basis of the classification and learns further patterns and regularities in the course of a learning transfer based thereon, the underlying classification is based on the predetermined trigger definitions.

According to embodiments, meta and / or context information on the classified tokens is provided in the form of the classification. This meta and / or context information is identified on the basis of the trigger in accordance with the trigger definitions and assigned to the corresponding token in the form of the class assignment. According to embodiments, the learning module can be configured to learn further patterns and regularities using this meta and / or context information.

Embodiments can have the advantage that the data records received from the source database are all stored in their original form in the document-oriented data model. This ensures that the full information content of these data records is retained. In addition, the data comprised by the data records stored in the document-oriented data model are provided in the form of the index. This index contains the corresponding data of the document-oriented data model in the form of tokens. According to embodiments, the index comprises all of the elementary data elements included in the document-oriented data model in the form of elementary tokens. According to embodiments, the index comprises data elements derived from the tokens. According to further embodiments, the index additionally comprises combinations of the elementary data elements in the form of token combinations that are comprised by the document-oriented data model. These tokens Combinations each include a combination of a plurality of elementary tokens. According to further embodiments, the index comprises token combinations up to a predetermined complexity. The complexity of a token combination is defined, for example, by the number and / or type of elementary tokens it comprises.

The tokens included in the index can be triggers according to the predetermined trigger definitions, for example. If a corresponding token is generated for the first time, for example in the course of processing an additional data record, it is identified as a trigger using one of the trigger definitions, added to the index and assigned to the corresponding trigger definition. If the learning module recognizes the same token, which the index defines as a trigger, within a further data record, the learning module accesses the trigger definition assigned to triggering in the index and assigns one or more tokens from a context environment following the corresponding trigger definition of the token stored as a trigger in the further data record to one or more classes of the first group of classes.

Furthermore, the index comprises a plurality of tokens which are each assigned to one or more classes of the first group of classes. According to embodiments, the assignment to the classes provides meta and / or context information on the corresponding tokens. The corresponding meta and / or context information can be used, for example, for processing the corresponding tokens and / or the data records comprising the corresponding tokens in the document-oriented data model. For example, the corresponding meta and / or context information is used in the course of a search query to identify relevant tokens and / or data records or in the course of a further method for machine learning that uses the index. Using the corresponding meta and / or context information, additional patterns and regularities can be learned in the course of a further learning transfer. This further method for machine learning is carried out, for example, by the learning module using the trigger definitions or a further learning module. For example, the further method for machine learning is an AI method that is executed by an AI module.

• • Zuordnen der verbleibenden zusätzlichen Token, für welche keine Zuordnung zu einer der Trigger-Definitionen und keine Klassenzuordnung aufgrund einer der Trigger-Definitionen erfolgt ist, zu der Auffangklasse.

According to embodiments, remaining tokens that are assigned neither to the trigger definition nor to one or more classes of the first group of classes are assigned to a collection class in the index of the source database for identifying the corresponding remaining tokens as unknown data, the assignment to the collection class being a Allocation to one of the trigger definitions as well as an allocation to one of the classes of the first group of classes excludes,
wherein the classifying further comprises:

• • Assignment of the remaining additional tokens, for which no assignment to one of the trigger definitions and no class assignment has been made based on one of the trigger definitions, to the collection class.

In this case, the index also includes tokens that do not fall under any of the predetermined trigger definitions. These tokens are neither triggers nor can they be assigned to the trigger classes defined by the trigger definitions. Rather, these tokens are unknown data that cannot be assigned and for which meta or context information is missing. These tokens are assigned to a trap class as unknown data. An assignment to the collection class excludes an assignment to one of the trigger definitions as well as an assignment to one of the classes of the first groups of classes.

Embodiments can have the advantage that, on the basis of the token assignments, it can be recognized in a simple form which tokens are unknown data and which tokens are known data, i.e. Triggers or classifiable data. Thus, for example, search queries can be defined in such a way that they only take into account known data. Additional learning algorithms can be configured, for example, in such a way that they work exclusively on known data. Thus, for example, a use of chance can be used in a decision-making or classification process, even if additional learning algorithms are used. In this case, the basis for all learning processes and / or KI processes is provided by the initially defined triggers, which are used to classify the data received from the database. According to embodiments, the predetermined trigger definitions provide a basis for supervised learning. By using the predetermined trigger definitions, undesirable developments in self-adhesive systems and / or in the processing of search queries that use the data stored in the database can be avoided.

Embodiments can furthermore have the advantage that additional data records that are added to the source database are each analyzed to determine which of the data they comprise is known data and which data is unknown data. In this context Known data means data that are known as triggers, for which meta or context information is available and / or for which meta or context information can be derived from the context of the data records using the trigger definitions. Data that are neither triggers nor data that can be classified using the trigger definitions are unknown data. Unknown data is assigned to the trap class.

Embodiments can have the advantage that a database system optimized for machine learning is used. The corresponding database system includes all the data on which machine learning is based, i.e. both trigger definitions used to classify data and the data that are processed using the trigger definitions. This enables continuous learning by the learning module, taking into account all of the data seen by the system or the learning module.

The source database stores all data records received in a document-oriented data model. A document-oriented data model means that the data model does not impose any structural requirements on the data to be saved. Rather, the data are stored in documents or data containers in the form in which they are received. In this sense, the data stored in the document-oriented data model are raw data. Raw data means that the data are stored in the form in which they are received, without additional data processing by the database management system, in particular no restructuring of the data. Embodiments can have the advantage that the entire information content of the received data can thus be (almost) completely retained without the assumptions of the database management system being included. Both the database management system and the learning module can access the original data at any time and take them into account in further processing.

An index is generated based on this data pool of raw data provided by the document-based data model. Structural information or contexts of meaning are only extracted from the data sets at this level. This structural information or context is taken into account in the form of class assignments of the indexed data.

For this purpose, the data records are broken down to an elementary level by tokenization, which takes into account the elementary components of the data records in the form of tokens. The tokens are assigned by the learning module as triggers to one of the trigger definitions or classified as using the trigger definitions. All tokens that are neither identified as triggers nor classified using one of the trigger definitions are assigned to the trap class as unknown data. For example, the learning module comprises a classifier and is configured to classify the tokens using the predetermined trigger definitions. The corresponding classification can, for example, be part of a pattern recognition in which a feature extraction is implemented by the tokenization. Based on this feature extraction, however, there is no feature reduction in the classic sense, since the complete database is indexed and thus each token is recognized as a trigger or assigned to a class, at least the capture class.

In addition, each token in the index is linked to one or more pointers that indicate in which data records the corresponding token occurs. This means that the raw data relevant to a token can be accessed at any time and this raw data can be used for evaluation with regard to this token.

As a result, the token assignments are differentiated according to known data, which represent reliable facts, and unknown data. Embodiments can have the advantage that the use of the structures and regularities determined by the learning module in the data records, which are reflected in the token assignments, is based on the use of the predetermined trigger definitions. Unknown data, on the other hand, are recorded as such and left out until they can also be classified and thus viewed as reliable facts. Such an additional classification can be implemented, for example, by additional trigger definitions. In particular, additional trigger definitions can be added to reduce the amount of tokens included in the collection class. The method thus enables learning and / or classification with reservations.

Embodiments can therefore have the advantage that they allow the learning module to work on the entire available database. In particular, they can have the advantage of enabling continuous learning which takes into account both additional data sets and data sets that have already been stored. Embodiments can therefore have the advantage that they are not restricted to an arbitrary subset being picked from an available total amount of data in order to use this to train. Rather, all of the data contained in the database are processed using the trigger definitions. Through additions to the trigger definitions, it can also be achieved, according to embodiments, that all tokens are either identified as triggers or classified using the (added) trigger definitions. If unknown data is excluded from search queries and / or further learning processes, this exclusion is not arbitrary, but based on the trigger definitions provided.

Embodiments can have the advantage that no random initialization is required, as is the case with known self-learning systems, e.g. neural networks. Rather, the initialization is based on the predetermined trigger definitions. Because of the random moment resulting from this random initialization, the decisions / classifications of a corresponding neural network are not transparent and cannot be traced. In contrast, embodiments can have the advantage of being completely deterministic.

Embodiments can have the advantage that an already trained system, e.g. the pre-trained learning module is retrained or trained further. Trigger definitions can be added, removed or changed. In this way, for example, the classes used in the classification can be added, removed or changed. If trigger definitions are added, removed or changed, then all assignments of tokens based on these to the corresponding trigger definitions or to one of the classes must be adapted accordingly. By adding trigger definitions, for example, new structures can be recorded that have not yet been shown. In this way, the factual knowledge in particular can be expanded subsequently, i.e. tokens assigned to the trap class can be assigned to other classes.

According to embodiments, the learning module implements a machine learning algorithm, the method not being restricted to a specific algorithm. According to embodiments, the machine learning algorithm comprises at least one classification algorithm for classifying tokens. Machine learning can be supervised or unsupervised learning. Machine learning can include classification and / or regression analysis. A learning algorithm tries to find a hypothesis or a mapping which assigns the (assumed) output value to each input value. If the output values to be assigned are present in a continuous distribution, the results of which can assume any quantitative values within a given range of values, this is generally referred to as a regression problem. If, on the other hand, the output values to be assigned are in discrete form or if the values are qualitative, this is generally referred to as a classification problem. According to embodiments, the machine learning relies on the classification of the indexed tokens. According to embodiments of the invention, the learning module comprises an algorithm specially developed for machine learning, such as, for example, without being limited thereto, a density-based multidimensional outlier detection (“local outlier detection”), a random Forrest algorithm, a neural network , a support vector machine, a naive Bayesian classifier or a feedback similar to the feedback of a linear or non-linear controller.

A multi-model database is understood here to mean a database which is configured to support a plurality of different data models. A multi-model database is thus configured to store, index and query data in more than one data model. Data models are, for example, relational, column-oriented, document-oriented, graph-based, key-value-based, etc. A database model defines the structure in which data is stored in a database system, i.e. the form in which the data is organized, stored and processed.

In the following, a database is understood to be a (typically large) amount of data that is managed in a computer system by a database management system (DBMS) according to specific criteria. The data is organized in a large number of data sets. In the following, a database management system or DBMS is understood to mean an electronic system for storing and retrieving data. The data are preferably stored consistently and permanently in the DBMS and are efficiently made available to various application programs and users in a needs-based form. A DBMS can typically contain one or more databases and manage the data records contained therein. The DBMS can preferably be a field-oriented DBMS, that is to say a DBMS that is configured to store parts of individual data records, so-called field values, in several different fields.

In the following, a data record is understood to be a coherent set of data provided to the database system, which is managed by the database management system as a coherent set of data. For example, a record includes a set content-related data. According to embodiments, data sets are in each case stored as coherent data sets in the document-oriented data model. For example, a single data set can represent a specific physical object, for example a natural person or a device. The person can be an employee, a patient, a customer, etc., for example. The device can be, for example, a production device, a computer device, a computer or network element or a transport device. The corresponding data record can contain a predefined set of attribute values for this person or device (e.g. name or pseudonym, age, height, weight, date of birth, ID numbers, security certificates, authentication codes, biometric data, identifiers, date of manufacture, date of commissioning, configuration data, and others). For example, a data record can represent a group of content-related (belonging to an object) data fields, e.g. B. Item number, item size, item color, item name or the like. The classes' name ', address'and' date of birth 'could, for example, form the logical structure of a data record for the object type “person”. In data processing, data is stored in the form of data records in databases, whereby they are the subject of the processing of computer programs and are generated, read, changed and deleted by these.

A "NoSQL" (English for Not only SQL) DBMS is a DBMS that follows a non-relational approach to data storage and does not require any fixed table schemes. NoSQL DBMSs include in particular document-oriented DBMSs such as Apache Jackrabbit, BaseX, CouchDB, IBM Notes, MongoDB, graph databases such as Neo4j, OrientDB, InfoGrid, HyperGraphDB, Core Data, DEX, AllegroGraph, and 4store, distributed ACID DBMSs such as MySQL Cluster, Key -Value databases such as Chordless, Google BigTable, GT.M, InterSystems Cache, Membase, Redis, sorted key-value memories, multivalue databases, object databases such as Db4o, ZODB, column-oriented databases and temporal databases such as Cortex DB.

An index is a data structure which accelerates a search for certain data values by a database management system. An index consists of a collection of pointers (references) that define an order relation to several “indexed” data values (stored in the index). For example, B + trees are used for this. Each indexed data value is linked to further pointers, which refer to data records in which the indexed data value found is contained and which represent the database for creating the index. Database management systems use indices to quickly identify the desired data records in response to a search query, in that the index is first searched along the pointers for a data value which is identical to a reference value contained in the search query. Without an index, the data values of a field managed by the DBMS would have to be searched sequentially, while a search using the index, e.g. of a B + tree, often only has logarithmic complexity.

The index also ranks the indexed data, i. Token, classes, whereby the corresponding data is linked with meta or context information. This meta or context information can be used in a search and / or in a machine learning process on the data in the database.

In the following, a field is an area on a logical or physical data carrier which is managed by a DBMS, which is assigned to a predefined field type and which is created and intended for storing a field value of a data record. A field is therefore an element for storing a field value of a data record as defined above. Fields of a data record are managed jointly by a DBMS.

A field value is a data value that is part of a data record and is stored in a field of the data record. A field value can consist of a single word, a single number, or a combination of several words and / or numbers and / or other data formats, with different embodiments of the invention varying degrees of flexibility with regard to the nature and combinability of data types within the same Include field values.

A "tokenizer" is a program logic that receives data, for example a field value, as input, analyzes the data, e.g. to recognize delimiters or other decomposition criteria and patterns, and then decomposes the data into one or more tokens as a result of the analysis Token returns. It is also possible that not all data will be returned as tokens. For example, a full-text indexer can recognize semantically insignificant stop words and filter them out so that they are not indexed. Alternatively, all data is returned as. To “tokenize” a data value means breaking the data value into several components according to a certain scheme. The components represent the tokens. For example, natural-language texts can be divided up using predefined separators, such as spaces, periods or commas Generated components (words) are used as tokens. According to embodiments, all tokens are used for indexing. It is also possible that some tokens are not used for indexing (e.g. stop words) or that the tokens are additionally processed prior to indexing (e.g. reducing words to the stem). In this case, for processing search queries from a client computer system to a server computer system, the search value is preferably processed in the same way by the client computer system or the server computer system to ensure that the search values of the search queries correspond to the tokens contained in the index.

For example, a class defines a category or a type to which a token belongs. The class therefore assigns meta or context information to the token, for example in the form of a property. For example, a class can represent a certain attribute of a physical object in the form of a token. For example, data records to be saved which contain employee attributes, which classes such as “name”, “pseudonym”, “ID number”; "Access certificate for room R", "Access certificate for device G", "Access certificate for building GB", "Age". Each token can be assigned to one or more classes. Furthermore, combinations of tokens as independent tokens can in turn be assigned to one or more further classes.

The data records received are stored using a document-oriented data model. For example, all field values of the stored data records are transferred as tokens into a multi-dimensional key / value store (key / value store) or key value databases. The tokens are assigned token types and stored in a form that meets the sixth normal form.

According to embodiments, the transaction time and the validity time of the data records are also stored bit-temporally. The transaction time indicates the point in time at which a change to a data object in the database occurs. The validity time specifies a point in time or time period in which a data object in the modeled image of the real world has the state described. If both the validity and the transaction time are relevant, one speaks of bitemporal.

A key-value data model enables storage, retrieval and management of associative data fields. Values are uniquely identified using a key.

In a document-oriented data model, also known as a document store, documents or data containers form the basic unit for storing data. A document-oriented data model enables document-oriented information, also known as semi-structured data, to be stored, retrieved and managed. Databases based on a document-oriented data model belong to the NoSQL databases and form a subclass of the key-value stores. In a key-value store, the data is considered to be inherently opaque to the database, while a document-oriented database relies on internal structures in the stored documents to extract metadata. The semi-structured data model is a database model in which there is no separation between the data and the schema and the scope of the structure used depends on the purpose of the database. Each document within the data model is addressed via a unique identifier.

A combination of the different database concepts enables data sets to be saved as documents or containers (document store) and additionally in the form of an index, e.g. of a key-value memory to be converted into the 6th normal form. This key-value memory represents the entire data volume in the document memory, while the original data records are retained.

According to embodiments, selections are carried out exclusively in the key-value memory in the redundancy-free sixth normal form. Only the result is read from the document storage container. According to embodiments, in addition to reading and writing rights in the data records, a selection right is also implemented on the key-value memory. This means that you can work on the index alone without having to read out the underlying data.

The proposed multi-model database thus provides a complete normalization of the entire scope of data in the sixth normal form in addition to a schematic data storage based on a document memory. Embodiments can have the advantage that the index contains data elements of the data records, i. Token, as a key and each of these keys is assigned one or more pointers as values which indicate in which data records and / or fields of the data records the corresponding key, i.e. Token / data value, is stored as a field value.

This index therefore forms over all fields of the data records and their contents, ie the field values, the entire database with all of them Records from, so that all queries are handled in the index and the schema-less stored data of the document-oriented data model are only used to output the search results. The small size of the index compared to the schema-less data enables quick queries in any query combination.

A computer or computer system is understood here to mean a device which processes data using programmable arithmetic rules. A program or program instructions is understood here, without restriction, to be any type of computer program which comprises machine-readable instructions for controlling a functionality of a computer. A computer or computer system can comprise a communication interface for connection to the network, wherein the network can be a private or public network, in particular the Internet or another communication network. Depending on the embodiment, this connection can also be established via a cellular network.

A computer system can be a stationary computer system, such as a personal computer (PC) or a client or server integrated in a client-server environment. Furthermore, a computer system can be, for example, a mobile telecommunications device, in particular a smartphone, a portable computer such as a laptop PC or palmtop PC, a tablet PC, a personal digital assistant or the like.

With a corresponding computer system, it can also be an object of the Internet of Things ("smart device"), e.g. be a so-called "wearable" or "wearable computer", i.e. portable electronic devices or portable computer systems which are arranged on the body of the user during use. Examples of wearables are smartwatches, i.e. Watches with computer functionality and connectivity, activity trackers, i.e. Device for recording and sending fitness and / or health-related data, smart glasses, i.e. Glasses, the inside of which serve as a screen, or articles of clothing in which electronic aids for communication are incorporated.

According to further embodiments, a computer system can be a computer system of a means of transport, such as a car, airplane, ship or train with an on-board computer.

According to further embodiments, a computer system can be a control computer of a smart home system, an access point or router of a local WLAN network, a multimedia device with computer functionality and connectivity, such as a smart TV, a control system of a locking system or a " smart device "or intelligent objects, ie an everyday object with information technology, which receives added value through sensor-based information processing and communication.

A memory is understood here to mean both volatile and non-volatile electronic memories or digital storage media.

A non-volatile memory is understood here as an electronic memory for the permanent storage of data. A non-volatile memory can be configured as a non-changeable memory, which is also referred to as read-only memory (ROM), or as a changeable memory, which is also referred to as non-volatile memory (NVM). In particular, it can be an EEPROM, for example a Flash EEPROM, referred to as Flash for short. A non-volatile memory is characterized by the fact that the data stored on it are retained even after the power supply has been switched off.

A volatile electronic memory is a memory for the temporary storage of data, which is characterized in that all data is lost after the power supply is switched off. In particular, this can be a volatile random access memory, which is also referred to as a random access memory (RAM), or a volatile main memory of the processor.

A processor is understood here and in the following to be a logic circuit that is used to execute program instructions. The logic circuit can be implemented on one or more discrete components, in particular on a chip. In particular, a processor is understood to mean a microprocessor or a microprocessor system made up of several processor cores and / or several microprocessors.

According to embodiments, if one of the additional tokens is not included in the index, the corresponding additional token is supplemented under its class assignments in the index, and if one of the class assignments of an additional token included in the index is not included in the index supplementing the corresponding class assignment with the corresponding additional token in the index and linking the corresponding additional token in the index with the pointer to the additional data set stored in the document-oriented data model.

For those tokens of the additional data set which the index already includes, but for which one or more class assignments determined using the additional data set are not yet taken into account by the index, the corresponding class assignments are added. In addition, the pointer to the additional data record is added to the index for these tokens.

Finally, only the pointer to the additional data record is added for those tokens of the additional data record which the index already includes with all their class assignments determined for the additional data record.

Embodiments can have the advantage that it can always be ensured that the index has all tokens comprised by the data records of the corresponding individual database. In addition, the index includes all class assignments found for all corresponding tokens. In addition, each of the tokens of the index is linked with a pointer to all data records of the corresponding individual database which contain the corresponding token.

For example, in the course of pre-training the learning mode, an initial set of predetermined trigger definitions is established. In the course of data acquisition, data records are received and stored in the document-based data model. The stored data records are tokenized and class assignments are determined for the resulting tokens using the initially set trigger definitions and an initial index is generated for the resulting tokens. According to embodiments, the initial index includes all triggers included in the trigger definitions as tokens. According to alternative embodiments, tokens defined as triggers by the trigger definitions are only added to the index on the condition that they are included in one of the data records.

An assignment of a token to a class using a predetermined trigger function is a fact secured by the corresponding predetermined trigger function. There is no token which is not a trigger and which are not detected by any of the trigger definitions it is because of such factual knowledge. Rather, the corresponding tokens are assigned to the trap class as unknown data. Embodiments can thus have the advantage that using initially established trigger definitions, new data can be converted into known data, i.e. Triggers or tokens classified using trigger definitions, and unknown data can be classified, i.e. Token assigned to the trap class.

According to embodiments, the combinations of second additional tokens with one or more of the identified triggers that have triggered a class assignment according to one of the trigger definitions are identified in the index as classified combinations and class assignments are only made for combinations of second additional tokens and one or more identified triggers which are not marked as classified combinations.

Embodiments can have the advantage that for all token combinations for which a class assignment has already been taken into account or for which a class assignment has already been carried out, are each identified in the index as already classified. It can thus be avoided that the same classifications are carried out again for token combinations which the learning module has already seen before and fully taken into account in the course of the classifications. The system can thus be designed to be significantly more efficient. For example, in addition to elementary tokens, the index includes all token combinations for which a classification has already taken place, i.e. all token combinations that are to be marked as classified. According to embodiments, the corresponding token combinations in the index are each provided with a flag which indicates whether the corresponding token combination is a classified token combination.

According to embodiments, upon tokenization of an additional data record, a comparison with all token combinations marked as already classified is first carried out. The classification is not repeated for these token combinations; rather, there is only a link with the pointer to the additional data record. According to embodiments, the corresponding pointer is also linked to all of the tokens included in the token combination in the index. According to embodiments, the comparison first takes place with the largest, ie most extensive, token combinations of the index. For all token combinations of the additional data record already recognized as classified, only the pointer to the corresponding data record is stored in the source database. According to embodiments, the corresponding pointer is also linked to all of the tokens included in the token combination in the index. In this way, the incorporation of the data of the additional data record into the index can be significantly accelerated. According to embodiments, a comparison with further tokens takes place successively Combination, whereby the size or scope of the other token combinations used gradually decreases. According to embodiments, only those further token combinations with a smaller size or scope are taken into account which, as part of a larger or more extensive token combination, a match was not found in the course of the comparison. Embodiments can have the advantage that for extensive token combinations which are recognized as already classified, there is no additional comparison for sub-combinations comprised by the corresponding token combination. Rather, a corresponding comparison only takes place if the corresponding sub-combination is included in the additional data record as an independent token combination, independent of the corresponding more extensive token combination.

• • Identifizieren von ein oder mehreren Trigger-Kombinationen, welche jeweils von zumindest einem der Datensätzen umfasst sind und ein Kombinationskriterium erfüllen,
• • für jede der identifizierten Trigger-Kombinationen Kombinieren der Trigger-Definitionen der Trigger der entsprechenden Trigger-Kombinationen zu ein oder mehreren zusätzlichen kombinierten Trigger-Definitionen,
• • Ergänzen der Mehrzahl von vorbestimmten Trigger-Definitionen des Lernmoduls durch die ein oder mehreren zusätzlichen kombinierten Trigger-Definitionen.

According to embodiments, the method further comprises:

• • Identification of one or more trigger combinations, which are each included in at least one of the data sets and meet a combination criterion,
• • for each of the identified trigger combinations combining the trigger definitions of the triggers of the corresponding trigger combinations to one or more additional combined trigger definitions,
• • Supplementing the plurality of predetermined trigger definitions of the learning module with the one or more additional combined trigger definitions.

Embodiments can have the advantage that, based on the triggers identified by the initial trigger definitions, additional triggers can be identified in the form of trigger combinations. Based on these identified trigger combinations, combined trigger definitions can be determined from the initial trigger definitions, with which the plurality of predetermined trigger definitions can be expanded.

According to embodiments, several token combinations, which are comprised by the same data record and fall under the combined trigger definition, are combined with one another and the resulting combination is identified in the index as a classified combination. Embodiments can have the advantage that on the basis of combined trigger definitions, token combinations in the index can be identified as classified combinations, thereby avoiding unnecessary repetitions of classifications of already classified token combinations.

According to embodiments, the combination criterion comprises a minimum frequency for the corresponding trigger combination to occur in the data records. Embodiments can have the advantage that corresponding trigger combinations are only used to form a combined trigger definition if the corresponding trigger combination occur in the data records with a minimum frequency. This can prevent additional combined trigger definitions from being formed due to the accidental occurrence of triggers with different trigger definitions in one and the same data set. Such a random occurrence is to be expected from a certain size and / or complexity of the data records, without it being possible to draw conclusions about an underlying relationship between the triggers. However, if the corresponding trigger combinations occur more frequently, a connection can be concluded from them.

According to embodiments, the minimum frequency defines an absolute frequency value of the occurrence in the data records. The corresponding minimum frequency can be a minimum value for the occurrence of the corresponding trigger combination in all data records. The occurrence of the corresponding trigger combination is added up across all data records. If the resulting sum is greater than or equal to the minimum value, this is fulfilled. Furthermore, the minimum frequency can be a minimum value for the occurrence in one of the data records. The occurrence of the corresponding trigger combination is summed up individually for the individual data records. If one of the resulting sums meets the minimum value, the minimum frequency is present. According to alternative embodiments, the minimum value must be fulfilled by a predetermined number of data records or a predetermined percentage of the data records. The corresponding predetermined percentage is either a percentage of all data records in the database or all data records which comprise the corresponding trigger combination. According to alternative embodiments, the minimum value must be fulfilled by all data records and / or by all data records which comprise the corresponding trigger combination. Furthermore, the corresponding minimum frequency can be a minimum value for an average frequency of occurrence of the corresponding trigger combination in all data records in the database or in all data records which comprise the corresponding trigger combination.

According to embodiments, the minimum frequency defines a relative frequency value of the occurrence in the data records. To Embodiments, the corresponding minimum frequency is dependent on the number of data records and / or the number of tokens and / or the size of the data comprised by the data records. For example, the frequency value determined by the minimum frequency increases with the number of data records and / or the number of tokens and / or the size of the data comprised by the data records.

According to embodiments, the minimum frequency stipulates a relative frequency value of the occurrence in the data records relative to the frequencies of occurrence of one or more of the triggers comprised by the corresponding trigger combination in the data records. For example, the relative frequency value is dependent on the occurrence of the trigger with the highest frequency of occurrence, the trigger with the lowest frequency of occurrence and / or an average value of the occurrence of all triggers of the corresponding trigger combination. Embodiments can have the advantage that, when a relative frequency value is taken into account, the frequency of occurrence of one or more of the triggers comprised by the corresponding trigger combination in the decision-making process whether to supplement an additional combined trigger definition based on the corresponding trigger combination is incorporated. As before in the case of the absolute frequency value, the frequency of occurrence of the corresponding triggers can refer to an occurrence of the corresponding trigger in all data records, to an average occurrence in all data records, to a most frequent occurrence in one of the data records and / or to a minimal occurrence in refer to one of the data sets.

Embodiments can have the advantage that the relative frequency value is selected to be higher, the higher the frequencies of occurrence of the one or more corresponding triggers comprised by the trigger combination. It can thus be avoided that a trigger definition is generated on the basis of a trigger combination, the occurrence of which is random, i.e. whose triggers happen to be included in the same data set, without this indicating a connection between the corresponding triggers.

According to embodiments, the combination criterion comprises one or more conditions at relative positions of the triggers of the corresponding trigger combination to one another within one of the data sets. Embodiments can have the advantage that a relative position of the triggers of the corresponding trigger combination within the data set is taken into account for the combination criterion. A corresponding relative position of data within data records results from or is dependent on contextual relationships. Corresponding contextual relationships can therefore be read from the relative position. Depending on the type of data comprised by the data sets, the relative position can be a relative position in a one-dimensional, i.e. sequential data structure, such as a text or voice file, a two-dimensional data structure such as an image file, or a higher-dimensional, for example three-dimensional or n-dimensional, data structure.

According to embodiments, the trigger definitions each include a definition of a trigger structure which defines positions relative to one another for one or more triggers included in the corresponding trigger definition and one or more tokens to be assigned to one of the classes according to the corresponding trigger definition.

Embodiments can have the advantage that a corresponding trigger definition using one or more triggers defines how one or more tokens are to be classified depending on a relative position of the corresponding tokens to the corresponding triggers. Depending on the type of data structures comprised by the data records, the corresponding relative position can be a relative position in a one-dimensional, two-dimensional or higher-dimensional, for example three-dimensional or n-dimensional, data space.

According to embodiments, the definitions of the relative positions include at least one of the following definitions: the one or more tokens to be assigned are arranged according to a trigger included in the corresponding trigger definition, the one or more tokens to be assigned are before a trigger included in the corresponding trigger definition arranged, the one or more tokens to be assigned are each arranged between triggers included in the corresponding trigger definition.

For example, a trigger can trigger a classification of previous data, e.g. "[Before1] [Trigger1]". In this case the occurrence of the trigger “Trigger1” triggers a classification of the preceding data “before1”. According to embodiments, the trigger itself is part of the classification, i.e. the combination "[before1] [Trigger1]" is classified. According to embodiments, the trigger “Trigger1”, if it is recognized, is assigned as a trigger to the corresponding trigger definition.

A trigger can, for example, trigger a classification of subsequent data, eg "[Trigger2] [after1]". In this case the occurrence of the trigger “Trigger2” triggers a classification of the subsequent data "after1". According to embodiments, the trigger itself is part of the classification, ie the combination “[Trigger2] [after1]” is classified. According to embodiments, the trigger “Trigger2”, if it is recognized, is assigned as a trigger to the corresponding trigger definition.

A trigger can, for example, trigger a classification of preceding and following data, e.g. "[Before2] [trigger3] [after2]". In this case, the occurrence of the “Trigger3” trigger triggers a classification of the preceding data “before2” and the following data “after2”. According to embodiments, the trigger itself is part of the classification, i.e. the combination "[before2] [trigger3] [after2]" is classified. According to embodiments, the trigger “Trigger3”, if it is recognized, is assigned as a trigger to the corresponding trigger definition.

A combination of two or more triggers can, for example, trigger a classification of preceding, following and data arranged between the triggers, e.g. "[Before3] [trigger4] [between1] [trigger5] [after3]". In this case, the occurrence of the combination of the triggers “Trigger4” and “Trigger5” triggers a classification of the preceding data “before3”, the following data “after3” and the data in between “in between 1”. According to embodiments, the triggers themselves are part of the classification, i.e. the entire combination "[before3] [trigger4] [between1] [trigger5] [after3]" is classified. According to embodiments, the triggers “Trigger4” and “Trigger5”, if it is recognized, are assigned as triggers to the corresponding trigger definition.

According to embodiments, a trigger combination can comprise any number of triggers, e.g. "[Before4] [Trigger6] [between2] [Trigger7] [...] [Trigger6 + N] [between2 + N] [Trigger6 + (N + 1)] [after4]". In this case, the occurrence of the combination of triggers “Trigger6” to “Trigger6 + (N + 1)” triggers a classification of the preceding data “before4”, the following data “after4” and the data in between “in between2” to “in between2 + N ". According to embodiments, the triggers themselves are part of the classification, i.e. the entire combination "[before4] [Trigger6] [between2] [Trigger7] [...] [Trigger6 + N] [between2 + N] [Trigger6 + (N + 1)] [after4]" is classified. According to embodiments, the triggers “Trigger6” to “Trigger6 + (N + 1)” are assigned as a trigger to the corresponding trigger definition when it is recognized.

In the case of a text file, an example trigger definition can look like this: "[Identity] [Trigger1 = may over] [Identity] [Trigger2 = and] [Identity]". The phrase “may over” is a first trigger [Trigger1] and the phrase “and” is a second trigger [Trigger2]. The structure therefore corresponds to a structure of the form [before] [Trigger1] [between] [Trigger2] [after]. In this case, preceding data [before] are classified as an identity, and intermediate data [between] and subsequent data [after] are classified as identities, respectively.

Another example is: "[Trigger = the customer bears the damage,] [condition]". The phrase “the customer bears the damage” is a trigger. The structure therefore corresponds to the structure [Trigger] [after]. In this case, the subsequent data [after] is classified as a condition.

Another example is: "[Trigger1 = between] [Identity] [Trigger2 = and] [Identity]". The phrase “between” is a first trigger [Trigger1] and the phrase “and” is a second trigger [Trigger2]. The structure therefore corresponds to the form [Trigger1] [between] [Trigger2] [after]. For example, if a data record includes the phrase: “The business relationship between the customer and the bank [...]”, the tokens “between” and “and” are identified as triggers. Using this trigger combination, the tokens in between are classified as “the customer” as an identity, as are the subsequent tokens “of the bank”. The triggers do not trigger any classification for the preceding tokens "The business relationship". Thus, the preceding tokens are classified as unknown data of the trap class.

According to embodiments, a trigger definition can stipulate that a token which is located within a radius around a specific trigger in an n-dimensional data space is to be assigned to a specific class. According to embodiments, in addition to the distance between the token and the trigger, it can also be decisive for the class assignment in which spatial direction the token is correspondingly rejected by the trigger. This can, for example, be defined by a vector which defines the relative position of the token to the trigger. According to embodiments, a trigger definition can stipulate that a token which is arranged within a plurality of radii around one trigger of a plurality of triggers is to be assigned to a specific class. Here, the n-dimensional areas delimited by the individual radii overlap and delimit an n-dimensional or lower-dimensional intersection area in the n-dimensional data space. A token which is part of this n-dimensional or lower-dimensional intersection is, for example, assigned to a specific class.

According to embodiments, a maximum trigger distance is defined for each of the triggers in accordance with the trigger definitions, which defines a maximum distance relative to the corresponding trigger to which a trigger effect of the trigger is limited.

Embodiments can have the advantage that the corresponding maximum distance is a radius around the corresponding trigger in an n-dimensional data space. In the case of a one-dimensional data space, the trigger effect is limited to the corresponding maximum trigger distance in front of and behind the corresponding trigger. In the case of a two-dimensional data space, the trigger effect is limited to a two-dimensional circular area around the corresponding trigger. In the case of a three-dimensional data space, the trigger effect is limited to a spherical volume around the corresponding trigger. In the case of an n-dimensional data space, the trigger effect is limited to a volume of an n-dimensional sphere around the corresponding trigger.

According to embodiments, the maximum distance can depend on the spatial direction and be set to be of different sizes in different spatial directions.

According to embodiments, the maximum trigger spacing is identical for all triggers. According to embodiments, the maximum trigger spacing is identical for a subset of the triggers. According to embodiments, the maximum trigger distance is determined individually for each trigger. The corresponding maximum trigger distance can be a distance in a specific unit, depending on the type of data. For example, a sequential sequence in time is a time interval measured in a time unit, such as milliseconds, seconds or minutes. For example, a one-dimensional, two-dimensional or three-dimensional spatial data structure is a spatial distance in a spatial unit, such as millimeters, centimeters, decimeters or meters. In the case of image or video data, the distance can be based on pixels or voxels, for example. A corresponding distance can thus be, for example, a number of pixels or a number of voxels.

According to embodiments, the distance is a logical distance. This can for example be based on elementary data elements, such as elementary characters. Thus, a corresponding distance can be a number of characters, for example. Furthermore, the corresponding spacing can be a number of elementary data elements composed of elements, such as a number of words. For example, the number is limited to a certain part of speech. Furthermore, the spacing can be limited by logical elements in the data structure, such as a punctuation mark and / or a trigger.

• • Ergänzen des vortrainierten Lernmoduls um ein oder mehrere zusätzliche Trigger-Definitionen, welche zusätzliche Trigger für ein Ersetzen von Zuordnungen von Token in dem Index zu der Auffangklasse durch Zuordnungen zu ein oder mehreren Klassen einer zweiten Gruppe von Klassen im Zuge eines Reklassifizierens definieren,
• • Reklassifizieren von ein oder mehreren der Auffangklasse zugeordneten Token in dem Index, welche die zusätzlichen Trigger-Definitionen als zusätzliche Trigger definierten, wobei das Reklassifizieren durch das Lernmodul ein Ersetzen der Zuordnung zu der Auffangklasse durch eine Zuordnung zu der entsprechen zusätzlichen Trigger-Definition umfasst, welche das entsprechende Token als zusätzlichen Trigger umfasst,
• • Verwenden der zusätzlichen Trigger zum Reklassifizieren von ein oder mehreren der Auffangklasse zugeordneten Token in dem Index zu ein oder mehreren Klassen der zweiten Gruppe von Klassen durch das Lernmodul, falls die entsprechenden der Auffangklasse zugeordneten Token von einem der Datensätze in einer Kombination mit ein oder mehreren der zusätzlichen Triggern umfasst sind und die entsprechenden zusätzlichen Trigger gemäß der entsprechenden zusätzlichen Trigger-Definition eine entsprechende Zuordnung zu den ein oder mehreren Klassen der zweiten Gruppe von Klassen triggern.

According to embodiments, the method further comprises:

• • Supplementing the pre-trained learning module with one or more additional trigger definitions which define additional triggers for the replacement of assignments of tokens in the index to the receiving class with assignments to one or more classes of a second group of classes in the course of a reclassification,
• Reclassifying one or more tokens assigned to the collection class in the index, which defined the additional trigger definitions as additional triggers, the reclassification by the learning module comprising replacing the assignment to the collection class with an assignment to the corresponding additional trigger definition, which includes the corresponding token as an additional trigger,
• Use of the additional triggers for reclassifying one or more tokens assigned to the collection class in the index to one or more classes of the second group of classes by the learning module if the corresponding tokens assigned to the collection class are from one of the data records in a combination with one or more the additional triggers are included and the corresponding additional triggers trigger a corresponding assignment to the one or more classes of the second group of classes in accordance with the corresponding additional trigger definition.

Embodiments can have the advantage that by adding additional trigger definitions to the learning module, the number of tokens that are assigned to the collection class can be reduced. Additional trigger definitions can be supplemented in a targeted manner in order to reclassify those tokens that are assigned to the trap class. Additional trigger definitions can therefore be supplemented as a function of the data records which the corresponding individual database comprises and the unknown data which they comprise.

For example, additional trigger definitions are added until all tokens of the trap class are reclassified. According to embodiments, corresponding additional trigger definitions are added according to predefined intervals. Corresponding predefined intervals are defined in time, for example, based on the number of tokens included in the collection class, the amount of data stored in the corresponding individual database and / or the amount of data added to the corresponding individual database since the last addition.

According to embodiments, the second group comprises classes different from the classes in the first group. Embodiments can have the advantage that additional classes are defined so that those tokens of the capture class can be classified for which the meta or context information corresponding to the classes of the first group cannot be used. Rather, additional meta or context information can be defined and used by the classes of the second group.

According to embodiments, one or more classes of the second group are each identical to one of the classes of the first group. Embodiments can have the advantage that the additional trigger definitions provide triggers which enable the tokens of the collection class to be assigned to classes of the first group of classes.

According to embodiments, the trigger definitions to be supplemented are each dependent, as supplements, on a trigger definition already included in the learning module. Embodiments can have the advantage that one or more of the supplementary trigger definitions are defined in the form of supplements to the trigger definitions already included in the learning module. The corresponding supplementary trigger definitions extend, for example, the trigger effect of already existing trigger definitions. According to embodiments, the supplementary trigger definitions form trigger definitions combined with the already existing trigger definitions.

According to embodiments, the additions are carried out repeatedly following a recursive scheme, the trigger definitions to be added to each recursion level each comprising additions to a trigger definition of a preceding recursion level so that the recursive additions form tree structures which each include one of the predetermined trigger definition as a root node .

Embodiments can have the advantage that the trigger effect of the existing trigger definitions can be successively expanded by a progressive recursion scheme until all tokens of the collection class are reclassified. The result of the corresponding additions to the already existing trigger functions can be tree structures, for example, which can then be followed by a classification of tokens.

According to embodiments, the additional trigger definitions to be supplemented are received by the learning module. Embodiments can have the advantage that the corresponding trigger definitions can be provided from outside, for example by an administrator. As a result, the relevant administrator always has the option of controlling, correcting and adding to the classification.

According to embodiments, an external fine adjustment can optionally or optionally take place, for example by an administrator. According to embodiments, additional trigger definitions from the class of the unknown data, i.e. the collecting class, tokens are extracted and assigned to existing classes and / or new classes are generated to which extracted tokens are assigned. For example, for the tokens classified as unknown data in the collection class, an administrator provides additional trigger definitions that are applied to the collection class, analogous to the initially provided trigger definitions. According to embodiments, the additional triggers are applied exclusively to the collection class and to data received in the future in accordance with the additional trigger definitions. According to embodiments, the use of an additional trigger can be implemented as an IF condition. For example, if another trigger has already been successfully applied to a data record, e.g. a trigger1, and the data set also includes data classified as unknown, where an additional trigger, e.g. a trigger2, applied according to one of the additional trigger definitions.

This fine adjustment can be repeated several times as a recursion. For example, the recursion is continued until the trap class no longer includes tokens, ie no more unknown data exists, or the number of tokens comprised by the trap class reaches and / or falls below a predefined threshold value, ie a predefined maximum number. The corresponding threshold value can be an absolute value which is independent of the number of tokens included in the index and the amount of data included in the corresponding individual database. Alternatively, it can be the corresponding threshold value is a relative value, which is dependent on the number of tokens included in the index and / or the amount of data included in the corresponding individual database

In this way, trigger trees or decision trees can be created behind the initially defined triggers or trigger definitions, the number of levels depending on the number of recursions N, e.g. the number of levels is N + 1. For example, each initial trigger or each initial trigger definition forms a root point of a corresponding trigger tree or decision tree. A decision tree is understood here to mean ordered, directed trees that serve to represent decision rules. If a data record includes an initial trigger, which means that part of the tokens in the data record can be classified without all tokens in the data record being able to be classified at the same time, it is checked whether the data record also includes a trigger of the first recursion. If the data record includes a trigger of the first recursion, whereby a further part of the tokens of the data record can be classified without simultaneously being able to classify all of the data of the data record, it is checked whether the data record also includes a trigger of the second recursion and so on away.

According to embodiments, the additional trigger definitions to be supplemented are created by the learning module, which comprises a statistical model, the statistical model being used for a statistical analysis of the tokens included in the collection classes and their occurrence in the data records, the result of the statistical analysis is used to create the additional trigger definitions to be added.

Embodiments can have the advantage that the learning module can independently create supplementary additional trigger definitions. For example, the optional or facultative fine adjustment described above takes place using the statistical model. The human action of an administrator described above can be replaced and / or improved by using a statistical model. According to embodiments, the statistical model identifies, e.g. by frequency analyzes and correlation analyzes, triggers within the unknown data, which are then applied to the tokens classified as unknown in analogy to the procedure described above. According to embodiments, a recursive procedure can also take place using the statistical model.

• • Empfang einer korrigierten Trigger-Definition zum Ersetzen einer der gespeicherten Trigger-Definitionen des Lernmoduls,
• • Ersetzen der entsprechenden gespeicherten Trigger-Definition durch die korrigierte Trigger-Definition,
• • Reklassifizieren der unter Verwendung der entsprechenden gespeicherten Trigger-Definition klassifizierten Token, wobei das Reklassifizieren unter Verwendung der korrigierten Trigger-Definition erfolgt.

According to embodiments, the method further comprises:

• • Receipt of a corrected trigger definition to replace one of the stored trigger definitions of the learning module,
• • Replacement of the corresponding stored trigger definition with the corrected trigger definition,
• Reclassification of the tokens classified using the corresponding stored trigger definition, the reclassification taking place using the corrected trigger definition.

According to embodiments, for example, an administrator can recognize errors in classified classes and correct them if necessary, for example by providing a corrected trigger definition on the basis of which tokens are reclassified. Embodiments can have the advantage that a correction of trigger definitions is made possible at any point in time of the method. For example, the trigger definitions can be checked after training the learning module. If correction trigger definitions are identified, correspondingly corrected trigger definitions are provided.

Embodiments can have the advantage that corrected trigger definitions can also be provided at a later point in time when incorrect classifications are recognized. Administrative intervention in the learning and classification process is therefore possible at any time. This means that errors in the learning system can be corrected without having to convert the entire model.

According to embodiments, the pointers with which the tokens are stored linked in the index each point to one or more of the field values in the stored data records.

Embodiments can have the advantage that a finer granularity can be achieved when determining the origin of tokens in the data records. Such a finer granularity also enables the relative relationships of the tokens within the data records to be broken down and taken into account in an analysis or other use of the index.

According to embodiments, the field values of the additional data record include text data, image data, audio data and / or video data. According to embodiments, the method can be used, for example, for signal processing, such as 1D audio recognition, 2D and 3D image processing, or ND data input from N sensors, etc. The method can be used, for example, for an analysis of stream data (bitstream or bitstream). A bit stream, also known as a bit stream, designates a sequence of bits that represent a flow of information, ie a serial or sequential signal. A bit stream is thus a sequence of bits of indefinite length in chronological order. A bit stream, for example, represents a data stream divided into logical structures, which can be divided into more fundamental small structures such as symbols of a fixed size, ie bits and bytes, and further into blocks and data packets of different protocols and formats.

According to embodiments, generating the tokens includes applying tokenization logic to the field values of the additional data record, which logic includes a full-text indexer that is configured to break down texts into words and to output the words as tokens. Embodiments can have the advantage that an effective tokenization of texts or text files can be implemented. The corresponding text files can be any text. For example, the corresponding text files can be measured value files or algorithms for controlling computers and / or technical systems. According to embodiments, the field values of the additional data record include full texts, the full texts comprising words formed from letters of one or more alphabets and / or one or more numbers.

Full-text indexing involves breaking down texts into individual words, with the individual words of a text field then being stored in an index assigned to this field. Full text indexing is only supported if the corresponding field is used to selectively store a certain data type, e.g. CHAR, VARCHAR or TEXT, is configured. For example, natural language text in JSON format can be stored in a field.

According to embodiments, generating the tokens includes applying tokenization logic to the field values of the additional data record, which logic includes a generic tokenizer which is configured to recognize data of different data types in the field values and to generate from these tokens in different data types. Embodiments can have the advantage that effective tokenization can be implemented for different types of data, such as text data, image data, audio data and / or video data.

• • Empfangen eines von einer dritten Einzeldatenbank als zusätzlicher Quelldatenbank erzeugten weiteren Ergänzungsdatensatzes durch die erste Einzeldatenbank über das Netzwerk, wobei der weitere Ergänzungsdatensatz in einem Index der zusätzlichen Quelldatenbank vorgenommenen Ergänzungen umfasst,
• • Integrieren des weiteren Ergänzungsdatensatzes in den Index der ersten Einzeldatenbank, wobei das Integrieren ein Ergänzen des entsprechenden Index umfasst, bei welchem
  • ◯ ein Abgleichen von Token des weiteren Ergänzungsdatensatzes mit dem Index der Quelldatenbank,
  • ◯ falls eines der Token des weiteren Ergänzungsdatensatzes nicht von dem Index der Quelldatenbank umfasst ist, Ergänzen des entsprechenden Tokens in dem Index der Quelldatenbank und Verknüpfen des entsprechenden Tokens in dem Index mit einem Zeiger auf den in einem dokumentenorientierten Datenmodell der zusätzlichen Quelldatenbank gespeicherten Datensatz, aus welchem das entsprechende Token erzeugt wurde,
  • ◯ falls eines der Token des weiteren Ergänzungsdatensatzes von dem Index der Quelldatenbank umfasst ist, Verknüpfen des entsprechenden Tokens in dem Index der Quelldatenbank mit dem Zeiger auf den in einem dokumentenorientierten Datenmodell der zusätzlichen Quelldatenbank gespeicherten Datensatz, aus welchem das entsprechende Token erzeugt wurde.

According to embodiments, the method further comprises:

• Receipt of a further supplementary data record generated by a third individual database as an additional source database by the first individual database via the network, the further supplementary data record comprising additions made in an index of the additional source database,
• • Integrating the further supplementary data set into the index of the first individual database, the integration comprising supplementing the corresponding index in which
  • ◯ a comparison of tokens of the further supplementary data set with the index of the source database,
  • ◯ if one of the tokens of the further supplementary data set is not included in the index of the source database, supplementing the corresponding token in the index of the source database and linking the corresponding token in the index with a pointer to the data set stored in a document-oriented data model of the additional source database to which the corresponding token was generated,
  • ◯ if one of the tokens of the additional supplementary data set is included in the index of the source database, linking the corresponding token in the index of the source database with the pointer to the data set stored in a document-oriented data model of the additional source database from which the corresponding token was generated.

Embodiments can have the advantage that the index of the source database can also be expanded by supplementary data records from other individual databases. The third individual database is, for example, an individual database which in turn receives supplementary data records from the source database. Changes to the indexes can thus be mutually exchanged between individual databases and uniform copies of a cross-database index can be implemented on a distributed database system, more precisely in the individual databases.

• • Empfangen einer Suchanfrage durch erste Einzeldatenbank, wobei die Suchanfrage einen oder mehrere Suchwerte beinhaltet,
• • Durchsuchen des Index der ersten Einzeldatenbank nach dem Suchwert,
• • Identifizieren ein oder mehrerer Token innerhalb des Index, welche mit einem oder mehreren der Suchwerte übereinstimmen,
• • Zurückgeben einer Antwort auf die Suchanfrage, wobei die Antwort zumindest umfasst: eine Angabe über die identifizierten Token, ein oder mehrere durch Analysieren von Zeigern, mit denen die identifizierte Token verknüpft sind, bestimmten Datensätze oder ein oder mehreren Referenzen auf die bestimmten Datensätze.

According to embodiments, the method further comprises:

• • Receipt of a search query by the first individual database, the search query containing one or more search values,
• • Searching the index of the first individual database for the search value,
• • Identify one or more tokens within the index that match one or more of the search values,
• Returning a response to the search query, the response comprising at least: an indication of the identified tokens, one or more specific data records or one or more references to the specific data records obtained by analyzing pointers with which the identified tokens are associated.

Embodiments can have the advantage that the index can be used for effective searches. The searches can be limited to information in the index, for example in how many or in which data records a certain search value occurs. According to embodiments, this requires index access authorization. Furthermore, a search on the data records of the local, i.e. The individual database executing the search takes place even though the data records are stored in their original form. For this purpose, according to embodiments, a read access authorization is necessary. In the case of data records that are stored in another individual database, these can be explicitly requested for reading. For this purpose, according to embodiments, read access authorization is again necessary.

In the case of the learning module, it can search for patterns and / or regularities within the data records using appropriate search queries.

According to embodiments, the index stores all tokens generated from the field values of the data records of a corresponding individual database in such a way that the index contains each token only once. Each token contains pointers to one or more of the data records from whose field values it was generated. If an index generated according to the invention is searched for a specific search value and a token stored in the index that is identical to the search value is identified as the result of the search, then this token uses pointers to refer to all data records that this token has at least once in at least contain one of their field values and which were used when the index was created. The data records, which represent a “hit” with regard to the search value, can be identified and returned very quickly using the references, without the need for a sequential search across all data records.

According to embodiments, the search value further comprises a class assignment and the identification of the token within the index further requires that the identified token has the same class assignment. Embodiments can have the advantage that class assignments and thus meta or context information indexed with the class assignments can be taken into account in the search queries.

According to embodiments, triggers are identified in the index with a flag. According to embodiments, the search value further includes an assignment to a trigger definition and / or a flag identifying a trigger, and the identification of the token within the index furthermore requires that the identified token is assigned to the same trigger definition and / or has the same flag.

According to embodiments, tokens which are assigned to the trap class are excluded from the search. Embodiments can have the advantage that the resulting search results have a high degree of reliability, since unknown data are excluded from the search.

• • Empfangen der Mehrzahl von vorbestimmten Trigger-Definitionen durch das Lernmodul, welche die Trigger für das Zuordnen von Token zu den Klassen der ersten Gruppe von Klassen definieren,
• • Speichern der empfangenen Mehrzahl von vorbestimmten Trigger-Definitionen durch das Lernmodul.

According to embodiments, the method further comprises pre-training the learning module. The pre-training includes:

• • Receipt by the learning module of the plurality of predetermined trigger definitions which define the triggers for assigning tokens to the classes of the first group of classes,
• • Storage of the received plurality of predetermined trigger definitions by the learning module.

According to embodiments, the predetermined trigger definitions initially define triggers that are used to structure or classify received data. According to embodiments, before data is loaded into the source database, the initial triggers are specifically defined, i.e. predetermined trigger definitions. If data is loaded, these initial triggers enable an initial classification according to known classes as well as unknown data, which are assigned to the collection class.

• • Extrahieren der Trigger aus der gespeicherten Mehrzahl von vorbestimmten Trigger-Definitionen durch das Lernmodul,
• • Zuordnen der initialen Token durch das Lernmodul jeweils zu der Trigger-Definition, welche das entsprechende initiale Token als Trigger umfasst,
• • Erzeugen des durchsuchbaren Index unter Verwendung der Mehrzahl von initialen Token durch das Multi-Modell-Datenbankmanagementsystem in dem weiteren Datenmodell, wobei der erzeugte Index die initialen Token umfasst, wobei jedes der initialen Token in dem Index jeweils eine Zuordnung aufweist zu der Trigger-Definition, welche das entsprechende initiale Token als Trigger umfasst.

According to embodiments, the pre-training further comprises:

• • Extraction of the triggers from the stored plurality of predetermined trigger definitions by the learning module,
• • Assignment of the initial tokens by the learning module to the trigger definition that includes the corresponding initial token as a trigger,
• • Generation of the searchable index using the plurality of initial tokens by the multi-model database management system in the further data model, wherein the generated index comprises the initial tokens, each of the initial tokens in the index each having an assignment to the trigger definition which comprises the corresponding initial token as a trigger.

After defining the trigger, data, e.g. Text data, audio data, image data, video data or N-dimensional data from N sensors, loaded into the source database and the triggers applied to automatically classify the data. This results in a fragmentation of the data into triggers, known classes, i.e. classes defined by the trigger definitions, and in unknown data.

Embodiments can have the advantage that the learning module can be effectively pretrained in this way on the basis of the predetermined trigger definitions.

These predetermined trigger definitions can serve as a basis for obtaining further trigger definitions, for example by combining trigger definitions. For example, there is an automatic learning phase of the database system or the learning module, which includes a combination of the initial triggers. In this way, the initially loaded triggers can be combined, as described above, based on the data comprised by the data records, and thus the number of trigger definitions available can be increased. In addition, token combinations that have already been classified can be identified. The purpose of this is to ensure that identical data that is later loaded into the source database does not have to be classified again, but is already marked as "known" in the system.

According to embodiments, generating one of the additional tokens comprises using one of the field values of the additional data set in its entirety as the corresponding additional token. It is entirely possible that the index also contains tokens from fields to which no tokenization is applied or whose content simply cannot be divided into individual tokens. According to embodiments, generating one of the additional tokens includes dividing one of the additional field values of the additional data record into a plurality of subfield values and using one of the subfield values as the corresponding additional token. Embodiments can have the advantage that the granularity of the data used or the tokenization can be adapted independently of the granularity of the fields.

According to embodiments, the index stores all tokens generated from the field values of the stored data records in such a way that the index contains each token precisely once for each of the token assignments of the corresponding token.

According to embodiments, the tokens, the class assignments and the assignment to the trigger definitions can be stored in the form of relations or equivalent structures. In the sense of relational database theory, a relation is a set of tuples. A tuple is a set of attribute values. An attribute describes a data type or a property assigned to one or more data. The number of attributes determines the degree, the number of tuples the cardinality of a relation.

According to embodiments, at least the document-based data model used by the multi-model database management system for storing the data records is a NoSQL data model. According to embodiments, the DBMS is a NoSQL DBMS. This can be advantageous because it has been found that NoSQL DBMS in particular, which often have a more flexible structure than classic SQL-based DBMSs. Due to the flexibility of their structure, NoSQL DBMSs are particularly suitable for managing and storing data records from which an index can be created in accordance with embodiments of the invention.

According to embodiments, the index has the structure of a tree, in particular a B ⁺ tree. Embodiments can have the advantage that a tree structure, in particular the structure of a B ⁺ tree, enables a particularly efficient and fast search for the tokens stored in the index. A B ⁺ tree is a data and / or index structure that is an extension of a B-tree. In a B ⁺ tree, the actual data elements are only stored in the leaf nodes, while the inner nodes only contain keys.

According to embodiments, several of the data records stored in a document-oriented data model each comprise a different number of fields. Embodiments can have the advantage that data sets of different sizes and structures or granularity can be processed.

According to embodiments, the fields each have a common, generic data format. Embodiments can have the advantage that since a large number of different data types can be stored in a specific field. A user or an application program that stores data records in the source database does not have to worry about the consistency and fit of data types. A high degree of flexibility with regard to the structure and the scope of the data records that can be managed and stored by the multi-model database management system is therefore offered.

According to embodiments, the learning module or the machine learning implemented by it is configured for data extraction, consistency checking, image recognition, speech recognition, voice control, device monitoring and / or autonomous device control. This can, for example, already consist in the classification of the tokens, with tokens assigned to the collection class as unknown data, for example, being viewed as an indication of a potential malfunction, for example. For example, this can be based on the index with the tokens and their meta or context information, which serve as the basis for an additional algorithm for machine learning applied to them. According to embodiments, for this purpose, the collection class is emptied by adding additional trigger definitions, so that meta or context information is provided for all tokens of the database system.

Data extraction can include, for example, recognizing and extracting a pattern in a text, image, audio or video file. This pattern can for example be defined by a trigger definition or it can be recorded in the classified data. A corresponding pattern can be, for example, a predetermined event recorded in the form of sensor values, for example a person in an effective area of a device.

A consistency check can include, for example, a consistency check in a text, image, audio or video file. In this case, it is checked, for example, whether the corresponding data include unknown and thus inconsistent data, whether data which differ greatly from the rest of the data or whether data that are explicitly predefined as inconsistent. A corresponding consistency check can serve, for example, to check for errors in control algorithms of devices, to detect malfunctions using measurement data of a function of a device, or to detect errors in text files, for example in the form of a spell check.

Image recognition can be used to recognize objects, events or features in image or video files. For example, context information on what is visually represented is recorded and / or displayed. This can include, for example, a visual representation of information, that is to say the addition of images or videos with computer-generated additional information or virtual objects by means of fading in / overlaying. Such a method is commonly referred to as augmented reality.

Speech recognition can be used to recognize speech in audio files or video files, for example for voice control or for converting speech into text form.

Pattern recognition in text, image, audio or video files can be used for device monitoring. In particular, occurring or impending malfunctions can be recognized in this way. This can serve safety and enables predictive maintenance of the corresponding device, since potential problems can be identified at an early stage. A corresponding text file is, for example, a data record with measured sensor values. Based on a device monitoring, an autonomous device control can also be implemented, for example an autonomous control of vehicles, robots or industrial plants.

A “device” here is generally understood to mean a technical device with sensors for acquiring status data of the device and a device computer system for logging the acquired status data. The device can also consist in the corresponding computer system with sensors. For example, the data records received are data records acquired by a device computer system using the sensors. Computer system for machine learning A device comprises, for example, a vehicle, a plant, such as a production plant, a processing plant, a conveyor system, an energy generation system, a heat generation system, a control system, a monitoring system, etc.

A “vehicle” is understood here to mean a mobile means of transport. Such a means of transport can be used, for example, to transport goods (freight transport), tools (machines or aids) or people (passenger transport). Vehicles in particular also include motorized means of transport. A vehicle can be, for example, a land vehicle, a watercraft and / or an aircraft. A land vehicle can be, for example: an automobile, such as a passenger car, bus or truck, a motor-driven two-wheeler such as a motorcycle, moped, scooter or motorbike, an agricultural tractor, forklift, golf cart, truck crane. In addition, a land vehicle can also be a rail-bound vehicle. For example, watercraft can be: a ship or Boat. Furthermore, an aircraft can be, for example: an airplane or a helicopter. A vehicle is also understood to mean, in particular, a motor vehicle.

According to embodiments, the device comprises at least one sensor for detecting status data of the device. The device status data is received by the device computer system from the at least one sensor. According to embodiments, the device comprises a plurality of sensors for acquiring status data of the device. Embodiments can have the advantage that the device's own sensor system can be used to detect the state of the device. The state of the device can be described, for example, by information on parameters of the current performance of the device, such as the mileage of a vehicle, consumption values, performance values, error messages, results of predefined test protocols and / or identifiers of components of the device.

Parameters of the current performance of a vehicle can be, for example, engine speed, speed, fuel consumption, exhaust gas values, transmission gear.

A “sensor” is understood here to mean an element for acquiring measurement data. Measurement data are data that qualitatively or quantitatively reproduce physical or chemical properties of a measurement object, such as the amount of heat, temperature, humidity, pressure, flow rate, sound field sizes, brightness, acceleration, pH value, ion strength, electrochemical potential and / or its material properties. Measurement data are recorded using physical or chemical effects and converted into an electronic signal that can be further processed electronically. Furthermore, measurement data can reproduce states and / or changes in state of electronic devices due to external influences and / or as a result of use by a user.

Sensors for acquiring status data in a vehicle can include, for example: crankshaft sensors, camshaft sensors, air mass sensors, air temperature sensors, cooling water temperature sensors, throttle sensors, knock sensors, transmission sensors, distance sensors, transmission sensors, level sensors, brake wear sensors, axle load sensors, steering angle sensors. These sensors record and monitor the driving behavior of the vehicle. Malfunctions can be recognized and identified from deviations from target values and / or the occurrence of specific patterns. In some cases, specific causes of errors, such as failed vehicle components, can also be identified. Sensors can also query the identifiers of electronic components installed in the vehicle in order to check their identity.

Embodiments include a source computer system for implementing a cross-database index on a distributed database system, which comprises a plurality of computer systems comprising the source computer system, each with an independent individual database, the computer systems with the individual databases being communicatively connected to one another via a network.

The computer systems also each include one or more processors, one or more data storage media on which the respective individual database is provided, a communication interface for communication via the network, and program logic.

The individual databases are each managed by a multi-model database management system, the individual databases each comprising a plurality of database-specific data records which are stored in a document-oriented first data model of the respective individual database, the stored data records each comprising one or more field values, the individual field values of the saved data records are each saved in a field.

The individual databases also each comprise a searchable first index which is stored in a second data model of the respective individual database, the index of the respective individual database comprising a plurality of tokens generated from the field values of the data records stored in the document-oriented data model of the corresponding individual database, the tokens is linked in the index with one or more pointers to one or more of the data records stored in the document-oriented data model of the corresponding individual database, from whose field values the corresponding token was generated.

• • Empfangen eines zusätzlichen Datensatzes durch die Einzeldatenbank des Quellcomputersystems als Quelldatenbank zum Ergänzen von Datensätze der Quelldatenbank,
• • Speichern des zusätzlichen Datensatzes, welcher ein oder mehrere zusätzliche Feldwerte umfasst, durch ein erstes Multi-Modell-Datenbankmanagementsystem der Quelldatenbank in einem dokumentenorientierten ersten Datenmodell der Quelldatenbank,
• • Anwenden einer Datenverarbeitungsfunktion der Quelldatenbank auf den zusätzlichen Datensatz, wobei die Datenverarbeitungsfunktion zumindest eine Indexierung des zusätzlichen Datensatzes zum Speichern in einem zweiten Datenmodell der Quelldatenbank umfasst, wobei die Indexierung umfasst:
  • ◯ Erzeugen von ein oder mehreren zusätzlichen Token aus den zusätzlichen Feldwerten durch die Quelldatenbank,
  • ◯ Ergänzen des Index der Quelldatenbank durch das erste Multi-Modell-Datenbankmanagementsystem unter Verwendung der zusätzlichen Token und eines Zeigers auf den zusätzlichen in dem dokumentenorientierten Datenmodell der Quelldatenbank gespeicherten Datensatz,
• • Erstellen eines aus dem Anwenden der Datenverarbeitungsfunktion resultierenden Ergänzungsdatensatzes, welcher die in dem Index der Quelldatenbank vorgenommenen Ergänzungen umfasst, zum Ergänzen zumindest eines zweiten Index zumindest einer weiteren Einzeldatenbank als Empfangsdatenbank, welche ein weiteres Computersystem des Datenbanksystems als Empfangscomputersystem umfassten,
• • Senden des Ergänzungsdatensatzes durch das Quellcomputersystem über das Netzwerk an das Empfangscomputersystem zur Integration in den zweiten Index der Empfangsdatenbank.

The program logic of the source computer system is configured to perform a method of implementing a cross-database index, the method comprising:

• • Receipt of an additional data set by the individual database of the source computer system as a source database to supplement data sets in the source database,
• • Storing the additional data record, which includes one or more additional field values, by a first multi-model database management system in the source database a document-oriented first data model of the source database,
• Applying a data processing function of the source database to the additional data record, the data processing function comprising at least one indexing of the additional data record for storage in a second data model of the source database, the indexing comprising:
  • ◯ Generation of one or more additional tokens from the additional field values by the source database,
  • ◯ Supplementing the index of the source database by the first multi-model database management system using the additional tokens and a pointer to the additional data set stored in the document-oriented data model of the source database,
• • Creation of a supplementary data set resulting from the use of the data processing function, which comprises the additions made in the index of the source database, for supplementing at least one second index of at least one further individual database as a receive database, which included a further computer system of the database system as a receive computer system,
• • Sending of the supplementary data set by the source computer system via the network to the receiving computer system for integration in the second index of the receiving database.

According to embodiments, the source computer system is configured to execute one or more of the aforementioned embodiments of the method for implementing a cross-database index on the distributed database system.

Embodiments further include a distributed database system for implementing a cross-database index on the distributed database system, which includes one of the plurality of computer systems each with an independent individual database, the computer systems with the individual databases being communicatively connected to one another via a network.

The computer systems also each include one or more processors, one or more data storage media on which the respective individual database is made available, a communication interface for communication via the network and program logic.

The individual databases are each managed by a multi-model database management system, the individual databases each comprising a plurality of database-specific data records which are stored in a document-oriented first data model of the respective individual database, the stored data records each comprising one or more field values, the individual field values of the saved data records are each saved in a field.

The individual databases also each comprise a searchable first index which is stored in a second data model of the respective individual database, the index of the respective individual database comprising a plurality of tokens generated from the field values of the data records stored in the document-oriented data model of the corresponding individual database, the tokens is linked in the index with one or more pointers to one or more of the data records stored in the document-oriented data model of the corresponding individual database, from whose field values the corresponding token was generated.

• • Empfangen eines zusätzlichen Datensatzes durch die Einzeldatenbank des Quellcomputersystems als Quelldatenbank zum Ergänzen von Datensätze der Quelldatenbank,
• • Speichern des zusätzlichen Datensatzes, welcher ein oder mehrere zusätzliche Feldwerte umfasst, durch ein erstes Multi-Modell-Datenbankmanagementsystem der Quelldatenbank in einem dokumentenorientierten ersten Datenmodell der Quelldatenbank,
• • Anwenden einer Datenverarbeitungsfunktion der Quelldatenbank auf den zusätzlichen Datensatz, wobei die Datenverarbeitungsfunktion zumindest eine Indexierung des zusätzlichen Datensatzes zum Speichern in einem zweiten Datenmodell der Quelldatenbank umfasst, wobei die Indexierung umfasst:
  • ◯ Erzeugen von ein oder mehreren zusätzlichen Token aus den zusätzlichen Feldwerten durch die Quelldatenbank,
  • ◯ Ergänzen des Index der Quelldatenbank durch das erste Multi-Modell-Datenbankmanagementsystem unter Verwendung der zusätzlichen Token und eines Zeigers auf den zusätzlichen in dem dokumentenorientierten Datenmodell der Quelldatenbank gespeicherten Datensatz,
• • Erstellen eines aus dem Anwenden der Datenverarbeitungsfunktion resultierenden Ergänzungsdatensatzes, welcher die in dem Index der Quelldatenbank vorgenommenen Ergänzungen umfasst, zum Ergänzen zumindest eines zweiten Index zumindest einer weiteren Einzeldatenbank als Empfangsdatenbank, welche ein weiteres Computersystem des Datenbanksystems als Empfangscomputersystem umfassten,
• • Senden des Ergänzungsdatensatzes durch das Quellcomputersystem über das Netzwerk an das Empfangscomputersystem zur Integration in den zweiten Index der Empfangsdatenbank.

The program logic is each configured to execute a method for implementing a cross-database index by the computer system executing the program logic of the plurality of computer systems, which method acts as a source computer system, the method comprising:

• • Receipt of an additional data set by the individual database of the source computer system as a source database to supplement data sets in the source database,
• • Saving the additional data record, which includes one or more additional field values, by a first multi-model database management system of the source database in a document-oriented first data model of the source database,
• Applying a data processing function of the source database to the additional data record, the data processing function comprising at least one indexing of the additional data record for storage in a second data model of the source database, the indexing comprising:
  • ◯ Generation of one or more additional tokens from the additional field values by the source database,
  • ◯ Supplementing the index of the source database by the first multi-model database management system using the additional tokens and a pointer to the additional data set stored in the document-oriented data model of the source database,
• • Creation of a supplementary data set resulting from the use of the data processing function, which comprises the additions made in the index of the source database, for supplementing at least one second index of at least one further individual database as a receive database, which included a further computer system of the database system as a receive computer system,
• • Sending of the supplementary data set by the source computer system via the network to the receiving computer system for integration in the second index of the receiving database.

According to embodiments, the distributed database system is configured to carry out one or more of the aforementioned embodiments of the method for implementing a cross-database index on the distributed database system.

• 1 ein schematisches Blockdiagramm einer Ausführungsform eines exemplarischen Computersystems,
• 2 schematische Diagramme einer Ausführungsform eines exemplarischen verteilten Datenbanksystems,
• 3 ein schematisches Diagramm einer Ausführungsform eines exemplarischen verteilten Datenbanksystems,
• 4 ein schematisches Flussdiagramm eines exemplarischen Verfahrens,
• 5 ein schematisches Flussdiagramm eines exemplarischen Verfahrens,
• 6 ein schematisches Flussdiagramm eines exemplarischen Verfahrens,
• 7 ein schematisches Flussdiagramm eines exemplarischen Verfahrens,
• 8 ein schematisches Blockdiagramm von Ausführungsformen exemplarischer Computersysteme,
• 9 ein schematisches Blockdiagramm einer exemplarischen Rechtevergabe,
• 10 ein schematisches Blockdiagramm einer exemplarischen Implementierung eines Systems zur Rechtevergabe,
• 11 ein schematisches Blockdiagramm einer Ausführungsform eines exemplarischen Computersystems,
• 12 ein schematisches Blockdiagramme einer exemplarischen Datenverarbeitung durch das Multi-Modell-Datenbankmanagementsystem,
• 13 ein schematisches Blockdiagramme einer exemplarischen Datenverarbeitung durch das Multi-Modell-Datenbankmanagementsystem,
• 14 ein schematisches Blockdiagramm von Ausführungsformen exemplarischer Computersysteme,
• 15 ein Flussdiagramm einer Ausführungsform eines exemplarischen Verfahrens,
• 16 ein Flussdiagramm einer Ausführungsform eines exemplarischen Verfahrens,
• 17 ein Flussdiagramm einer Ausführungsform eines exemplarischen Verfahrens und
• 18 ein Flussdiagramm einer Ausführungsform eines exemplarischen Verfahrens.

In the following, embodiments of the invention are explained in more detail with reference to the drawings. Show it:

• 1 a schematic block diagram of an embodiment of an exemplary computer system;
• 2 schematic diagrams of an embodiment of an exemplary distributed database system,
• 3 a schematic diagram of one embodiment of an exemplary distributed database system;
• 4th a schematic flow diagram of an exemplary method,
• 5 a schematic flow diagram of an exemplary method,
• 6th a schematic flow diagram of an exemplary method,
• 7th a schematic flow diagram of an exemplary method,
• 8th a schematic block diagram of embodiments of exemplary computer systems;
• 9 a schematic block diagram of an exemplary rights allocation,
• 10 a schematic block diagram of an exemplary implementation of a system for granting rights,
• 11 a schematic block diagram of an embodiment of an exemplary computer system;
• 12 a schematic block diagram of an exemplary data processing by the multi-model database management system,
• 13 a schematic block diagram of an exemplary data processing by the multi-model database management system,
• 14th a schematic block diagram of embodiments of exemplary computer systems;
• 15th a flow diagram of an embodiment of an exemplary method,
• 16 a flow diagram of an embodiment of an exemplary method,
• 17th FIG. 3 is a flow diagram of an embodiment of an exemplary method and FIG
• 18th Figure 6 is a flow diagram of one embodiment of an exemplary method.

Elements of the following embodiments that correspond to one another are identified by the same reference symbols.

1 Figure 3 shows a block diagram of one embodiment of an exemplary computer system 100 , which is a single database 104 to implement a cross-database index on a distributed database system 170 includes. The distributed database system 170 comprises a plurality of individual databases 104 each on a computer system 100 are implemented. The computer system 100 also includes a multi-model database management system (MM-DBMS) 118 , which is the, possibly structured, storage of data in the at least one individual database 104 manages and all read and write access to the individual database 104 controlled. The MM-DBMS 118 supports at least two data models 106 , 110 in which data in the individual database 104 get saved. The database model defines the form in which the relevant data is organized, saved and processed. With one or both data models 106 , 110 it is about NoSQL data models. In this respect, it is the MM-DBMS 118 a NoSQL DBMS. The first data model 106 is a document-based data model in which a plurality of data sets DS1 , DS2 , DS3 are stored. Every record DS1 , DS2 , DS3 is saved in a document or a data container. The records DS1 , DS2 , DS3 itself is saved from the document-based data model 106 no specific structure specified. Thus, the records DS1 , DS2 , DS3 are saved with the internal structure with which the data records DS1 , DS2 , DS3 from the individual database 104 be received. In this respect, it is the document-based Data model 106 stored records DS1 , DS2 , DS3 around raw data. The records DS1 , DS2 , DS3 can for example include text data, image data, audio data and / or video data. The records DS1 , DS2 , DS3 each include at least one field F1 , .. ., F8 , with field values. The records DS1 , DS2 , DS3 can already have an internal structure with a plurality of fields F1 , ..., F8 when they are saved. Then include the appropriate records DS1 , DS2 , DS3 each a plurality of field F1 , ..., F8 . Assign the records DS1 , DS2 , DS3 even when they are received, they contain, in stored form, exactly one field, for example, which contains the entire data volume of the corresponding data record DS1 , DS2 , DS3 includes. The fields F1 , ..., F8 each include one or more field values. Each of the field values of a record DS1 , DS2 , DS3 is stored in a corresponding field, a kind of data container. Every field F1 , ..., F8 can be assigned to a field type. The fields F1 , ..., F8 different or all of the same field type. The composition of the field values of the individual records DS1 , DS2 , DS3 can differ in terms of their field types. It is also possible that individual data records do not contain any fields of a certain field type. In other embodiments (not shown here) mandatory field types can also be defined, ie each document comprises a field for each mandatory field type and optionally comprises one or more further fields for optional field types. The data of the data records are then stored in fields of the field type intended for them, ie for example text data in one or more text fields, image data in one or more image fields, audio data in one or more audio fields and / or video data in one or more video fields.

The computer system also includes 100 a tokenizer 120 to generate tokens 109 . According to alternative embodiments, the MM-DBMS 118 the tokenizer 122 include.

The computer system 100 , e.g. the MM-DBMS 118 , has a built-in program logic or data processing function which is used to generate an index 112 configured. The corresponding index 112 is used in another data model 110 provided in which the complete data of the datasets DS1 , DS2 , DS3 or a data volume derived from these is stored in a restructured, redundancy-free form. To generate the index 112 is on the tokenizer 122 which is configured to read the field values of the fields F1 , ..., F8 that in the document-based data model 106 saved records 106 tokenize. The resulting tokens 109 also be identical to a field value of a field or a data record, if no further breakdown into tokens 109 is possible or useful. According to embodiments, the tokenization can also take place in stages, so that an ever finer breakdown takes place. The resulting index 112 in this case token 109 include those from other tokens 109 are composed. The program logic or data processing function can be used on all computer systems 100 or for all individual databases 104 of the distributed database system 170 be identical. Alternatively, the program logic or data processing function can be between computer systems 100 or individual databases 104 of the distributed database system 170 also differentiate, for example depending on the type or the content of the data to be stored by the respective individual database.

Preferably all or at least most of the field values of all data records are used DS1 , DS2 , DS3 the individual database 104 tokenized, so an extensive amount of tokens 109 arises. Depending on the type of data in the individual field values, the tokens 109 a mixture of numbers, letters, images or image segments, audio files or audio elements or other data structures, in particular sensor data from one or more sensors. Each of the generated tokens 109 is in the index 112 linked to a pointer, the pointer pointing to the data record or field from which the token 109 comes from.

In the course of generating the index 112 becomes from the amount of tokens 109 a non-redundant, unique token set is formed, in which each of the tokens 109 occurs at most once. All tokens are preferably stored 109 the non-redundant token set in the index 112 so that the token 109 sorted according to a sorting criterion and saved in sorted form in the index structure. The sorting can take place, for example, using the alphabet for alphanumeric data or other sorting criteria adapted to the data. Since the tokens 109 in the index 112 are preferably stored in sorted form, and are furthermore preferably stored in a tree structure, it is possible very quickly to obtain a specific token 109 within the index 112 to identify and then the references of this identified token 109 on one or more records DS1 , DS2 , DS3 to be used to very quickly identify those records that contain a particular token 109 contain. So it is not necessary to have all records DS1 , DS2 , DS3 the individual database 104 to browse sequentially.

Receives the individual database 104 an additional data record as the source database, e.g. DS3 , becomes this in the first data model 106 stored, tokenized using the tokenizer and the resulting tokens are used to create the index 112 in the second data model 110 to complete. Furthermore, a supplementary data record 130 which is created in the index 112 the source database 104 The additions made include and to add further indices to further individual databases of the distributed database system 170 serves as receiving databases. The supplementary data set created in this way 130 is made using a communication interface 126 of the computer system 100 via a network to the receiving databases of the distributed database system 170 Posted.

2A Figure 12 is a diagram of one embodiment of an exemplary distributed database system 170 . The distributed database system 170 includes a plurality of nodes K ₁ , ..., K ₆ . These knots K ₁ , ..., K ₆ are used by computer systems 100 , 200 formed, which each comprise a single database of the distributed database system. Each of the knots K ₁ , .. ., K ₆ aggregates individual data D ₁ , .. ., D ₆ and creates an index from these I ₁ , ..., I ₆ . Becomes an index I ₁ , .. ., I ₆ supplemented as a result of an additional received data record, the corresponding node is created K ₁ , ..., K ₆ , e.g. knot K ₁ , a supplementary data set E (I ₁ ) , which is the additions to the index I ₁ and sends it to the other accounts K ₂ , ..., K ₆ of the distributed database system 170 , e.g. knot K ₂ . Likewise create the other nodes K ₂ , ..., K ₆ of the distributed database system 170 when adding to your index, e.g. I ₂ , Supplementary data sets E (I ₂ ) and send them to the other nodes, such as nodes K ₂ . As a result of this exchange of supplementary data records E (I ₁ ) , E (I ₂ ) become all index I ₁ , .. ., I ₆ of the database system 170 synchronized so that the in 2 B state shown results in which all nodes K ₁ , .. ., K ₆ via a complete cross-database index I ( I ₁ , ..., 16 ) of the distributed database system 170 feature.

In 3 is a plurality of nodes K ₁ , ..., K _N of a distributed database system 170 shown, which each have a single database 1 , ..., N, include. As a result of the 2 synchronization methods illustrated include each of the individual databases 1 , ..., N a complete cross-database index of the distributed database system 170 , which contains the information of all individual indices I ₁ , ..., I _N includes.

4th shows a flow chart of the addition of the indexes. In step a) a first node K ₁ the distributed database 170 provided which has a data processing function F ₁ includes. In step b) the node receives K ₁ Data D ₁ . From this data D ₁ is done in step c) using the data processing function F ₁ an index I ₁ generated or an existing index I ₁ added. In step d) the resulting index I ₁ or one from the addition of the existing index I ₁ resulting supplementary data record from the node K ₁ to at least one other node K ₂ the distributed database 170 Posted. The other knot K ₂ also includes data D ₂ , from which using the data processing function F ₂ an index I ₂ is produced. The index I ₂ is about the received index information I ₁ of the first node K ₁ added so that the resulting index of the node K ₁ is a combined index I (I) which is the index information of both indexes I ₁ , I ₂ both nodes K ₁ , K ₂ combined with each other.

5 Figure 12 shows a flow diagram of an exemplary method for implementing a cross-database index on a distributed database system. In block 300 an additional data record is received by a first individual database of the distributed database system as the source database. This additional data record, which for example includes one or more additional field values, is used to supplement the source database. In block 302 the additional data record is stored by a multi-model database management system of the source database in a document-oriented first data model of the source database. In addition, a data processing function of the source database is applied to the additional data record, which includes at least indexing of the additional data record for storage in a second data model of the source database. In block 304 one or more additional tokens are generated from the additional data record. In block 306 the index of the source database is supplemented by the first multi-model database management system using the additional tokens and a pointer to the additional data record stored in the document-oriented data model of the source database. For example, the additional tokens are compared with the index of the source database. If one of the additional tokens is not included in the index of the source database, the corresponding additional token is added to the index of the source database and linked to the pointer to the additional data record stored in the document-oriented data model of the source database. If one of the additional tokens is already included in the index of the source database, the corresponding additional token in the index of the source database is linked to the pointer to the additional data record stored in the document-oriented data model of the source database.

In block 308 a supplementary data record resulting from the application of the data processing function is created. This Supplementary data set comprises the additions made to the index of the source database and is used to supplement at least one second index of at least one second individual database of the distributed database system as a receive database. In block 310 the supplementary data record is sent via a network to the second individual database for integration into the second index of the receiving database.

6th FIG. 8 shows a flow diagram of an exemplary method for integrating a supplementary data record into the index of a receive database. In block 400 the receiving database receives the supplementary data record from a source database. The supplementary data record includes additions made to the index of the source database. In block 402 the received supplementary data record is integrated into the index of the receive database. The integration includes adding to the index of the receiving database. For example, tokens of the supplementary data record are compared with the index of the source database. If one of the tokens of the supplementary data set is not included in the index of the source database, the corresponding token is supplemented in the index of the source database and in the index with a pointer to the data set stored in a document-oriented data model of the source database from which the corresponding token was generated , connected. If one of the tokens of the supplementary data record is already included in the index of the source database, the corresponding token is linked in the index of the source database with the pointer to the data record stored in a document-oriented data model of the additional source database from which the corresponding token was generated.

7th FIG. 3 shows a flow diagram of an embodiment of an exemplary method for performing a search on one of the individual databases. In block 500 a search query is received that contains a search value. block 502 the index is searched for the search value and in block 504 a token is identified within the index which is identical to the search value and / or which falls below the search scope defined by the search value. In block 506 If the search so requires, pointers with which the identified token or tokens are associated are analyzed. In this way, one or more of the data records can be determined which contain one or more field values from which the indexed token was generated. In block 508 an answer to the search query is returned. This response includes, for example: an indication of the identified tokens, one or more specific data records or one or more references to the specific data records by analyzing pointers with which the identified tokens are linked. If, for example, a search is only made for which tokens meet a certain criterion determined by the search value, it is sufficient to return the tokens found without further analysis of pointers or data records. If, according to the search, the data records that comprise the tokens found are also to be identified, the pointers are analyzed, but access to the underlying data records is not necessary. Alternatively, the underlying data records can also be queried according to the search.

8th Figure 3 shows a schematic block diagram of one embodiment of an exemplary source computer system 100 as well as an exemplary receiving computer system 200 of a distributed database system 170 . Both computer systems 100 , 200 each includes at least one processor 114 , 214 which program instructions 116 , 216 executes. By executing the program instructions 116 , 216 For example, the method described above for implementing a cross-database index on a distributed database system is used 170 implemented. For example, the processors lead 114 , 214 each a multi-model database management system 118 , 218 as well as a tokenizer 122 out. The computer systems also include 100 each in a memory 102 , 202 a single database 104 , 204 , which from the respective multi-model database management system 118 , 218 is managed. The databases 104 , 204 each comprise a first data model 106 , 206 , e.g. a document-oriented data model in which data records 108 , 208 get saved. The databases also include 104 , 204 a second data model each 110 , 210 with an index 112 , 212 all in the records 108 , 109 stored data. To the indexes 112 , 212 with different data sets 108 , 109 to be able to synchronize and thus a cross-database index on the distributed database system 170 to implement. For example, used on the source computer system 100 a supplementary data set 130 generated what additions to the index 112 includes. To these additions to the index 112 Both computer systems include being able to also make them available to the receiving computer system 100 , 200 one communication interface each 126 , 226 through which they have a network 180 communicate with each other. The source computer system 100 sends the supplementary data record 130 over the network 180 to the receiving computer system 200 , which is made using the supplementary data set received 130 its index 212 with the index 112 synchronized.

9 shows a block diagram of the allocation of rights by means of the transfer of access rights via a chain of entities, such as individual databases or their users. One record 108 with user data 160 is received through a single database. in the During the creation of the data set 108 or storage in the individual database, access certificates are automatically created in the background 162 , 164 , 166 that are assigned to a first entity are stored in corresponding fields of the data record. The first entity is, for example, the creator of the data record, such as the corresponding individual database or a user of the same. The access certificates 162, 164, 166 are configured for different types of access to the data record. The access certificate 162 is for read access, the access certificate 164 for write access and the access certificate 166 for access to tokens of an index, which are generated from the corresponding data record.

If a further entity is also to be granted access rights to the index, for example a second individual database or a user of the second individual database, an in an ID database 172 the index access certificate of the other entity 166 which index accesses in relation to the data record 108 allows assigned. This is done, for example, by linking this certificate 166 with a user certificate 168 the other entity in the ID database 172 is stored, for example within the same access authorization chain object.

In the same way, a third entity can also have access rights with regard to the data record 108 be assigned. The index access certificate is used for this 166 also assigned to this further entity. This is done, for example, by linking this certificate 166 with a user certificate 174 the third entity in the ID database 172 is stored, for example within the same access authorization chain object. The user certificate 174 either to the user certificate 168 or directly to the index access certificate 166 be appended.

Should there be access to tokens in an index, which from the data record 108 was generated, it is first checked whether the owner of the index access certificate itself 166 or one of the ones with the index access certificate 166 linked user certificates 168 , 174 can be proven. If this succeeds, access is granted, otherwise access is denied.

The chain of user certificates documented in the access authorization chain objects documents the transfer of access rights over a sequence of several users. The sequence can consist of a simple stringing together of user certificates, the position of the user certificates within the chain objects representing the time series of the transmissions, for example. Optionally, according to some embodiments, the chain of user certificates within an authorization chain object can be generated in that the ID database uses the private keys assigned to the individual user certificates so that the last user certificate in the chain is the new user added to it -Certificate signed so that a certificate chain check is also possible within the chain of user certificates of the individual authorization chain objects.

10 Figure 3 shows a block diagram of the distributed database system 170 . This includes, for example, a source computer system 100 and a receiving computer system 200 . The source computer system 100 sends over a network 180 a supplementary data set 130 to the receiving computer system 200 to add one from the receive database 204 of the receiving computer system 200 included index so that it matches an index of the source database 104 of the source computer system 100 is synchronized.

To access the index of the source database 104 must be the source computer system 100 or a user of the same access rights proof. These are, for example, compared with one in the source ID database 172 stored access authorization chain objects checked. If the access rights exist, the index can be accessed, for example, and one supplementary data record can be used 130 to be created. This also applies to the receiving computer system 200 to supplement the index of the receive database 204 is authorized, a corresponding access authorization chain object is generated for this or a user thereof and is transferred to the source ID database 174 transfer. Thus the receiving computer system can do the same 200 prove his access rights. According to embodiments is an authorization of the receiving computer system 200 or the receive database 204 a necessary prerequisite for sending the supplementary data record 130 with index information to the receiving database 204 . According to alternative embodiments, a central ID database for several or all computer systems 100 , 200 of the distributed database system 170 be provided.

11 Figure 3 shows a block diagram of an expanded embodiment of the exemplary computer system 100 out 1 which is configured for machine learning. The computer system 100 out 11 also includes a learning module 120 for processing the in the individual database 104 stored data. The learning module 120 includes, for example, the tokenizer 120 to generate tokens 109 and also trigger definitions 123 which triggers for a classification of tokens 109 define, and / or a classifier 124 which the token 109 using the trigger Definitions 123 classified. According to embodiments, the learning module comprises 120 also a statistical model 125 . The statistical model 125 can be configured to acquire trigger combinations and create combined trigger definitions, create additional trigger definitions and / or create corrected trigger definitions. According to alternative embodiments, the MM-DBMS 118 the tokenizer 122 include and / or on one of the learning module 120 provided tokenizer 122 access. According to alternative embodiments, the trigger definitions 123 also in the individual database 104 be saved.

The MM-DBMS 118 and / or the learning module 120 have a built-in program logic that is used to generate an index 112 configured.

When generating the index 112 become tokens 109 in the index 112 which of one of the trigger definitions 123 are included as triggers, each of the corresponding trigger definition 123 assigned. Furthermore, tokens 109 in the index 112 which from one of the records DS1 , DS2 , DS3 in a combination with one or more of the identified triggers according to one of the trigger definitions 123 are included, each assigned to one or more classes. The corresponding class assignments provide meta or context information for the corresponding tokens 109 ready. Finally, the remaining tokens are 109 in the index 112 which can be found using the trigger definitions 123 neither identify as trigger nor assign to a class as, assigned to a trap class to identify as unknown data. An assignment to the trap class includes an assignment to one of the trigger definitions 123 as well as an assignment to one of the classes according to the trigger definitions 123 out. The assignments described above are made using the classifier, for example 124 of the learning module 120 .

In the course of generating the index 112 becomes from the amount of tokens 109 a non-redundant, unique token set is formed, in which each of the tokens 109 occurs only once. Even if a token 109 with a certain value and a certain class assignment multiple times in the single database 104 or in the data model 106 occurs, for example, it is only once with this class assignment in the non-redundant token set and in the index 112 saved.

12 FIG. 10 shows a schematic block diagram of exemplary data processing by the multi-model database management system and the learning module. There will be a predetermined trigger definition 123 in the form "[First name] [Last name] [Trigger1 = lives in] [Street] [Trigger2 = in] [City]". This trigger definition 123 defines two triggers, ie a first trigger "lives in" and a second trigger "in". In addition, the trigger definition defines that a token immediately preceding the first trigger is a surname, while a token immediately preceding the surname is a first name. The trigger definition also defines that a token arranged between the two triggers is a street and that a token immediately following the second trigger is a city.

In a document-based data model 106 a database are two documents 108 saved. Any document 108 each includes one data set DS1 , DS2 . For example, it concerns the data records DS1 , DS2 one text file each. The first record DS1 includes, for example, the sentence: "Sample first name_1 Sample last name_1 lives in Musterstrasse_1 in Musterstadt_1". This set is converted into tokens by means of a tokenizer 109 disassembled: "Musterforname_1", "Mustername_1", "lives in the", "Musterstr._1", "in", "Musterstadt_1".

The two tokens "lives in" and "in" are used as triggers according to the trigger definition 123 identified. Using the identified triggers and the trigger definition 123 will be the remaining tokens 109 each of the classes defined by the trigger definition 111 assigned. The token “Musterfirename_1” is assigned to the class “First name”, the token “Musterlastname_1” to the class “Last name”, the token “Musterstrasse_1” to the class “Street” and the token “Musterstadt_1” to the class “City”. The tokens identified as triggers, like the tokens classified using these triggers, are stored in an index in a second data model 110 saved. The triggers are in the form of a trigger assignment 117 the trigger definition 123 assigned. The remaining tokens 109 are each in the form of a class assignment 113 one of the by the trigger definition 123 saved assigned to defined classes. In addition, all triggers and classified tokens are in the second data model 110 with a pointer 115 to their location in the first data model, ie DS1 , connected.

The database receives a second record DS2 of the form: "Musterforname_2, Musterstr._2 in Musterstadt_2", this sentence is broken down into tokens by means of a tokenizer: "Musterforname_2", "Musterlastname_2", "lives in the", "Musterstr._2", "in", " Musterstadt_2 ”and in a redundancy-free form also in the second data model 110 saved.

The two tokens "lives in" and "in" are used as triggers according to the trigger definition 123 identified. Because these two triggers are the trigger definition 123 are already included in the index, they are not included again in the second data model 110 saved. It just becomes a pointer to the second record DS2 added. Using the identified triggers and the trigger definition 123 will be the remaining tokens 109 of the data set DS2 each of the classes defined by the trigger definition 111 assigned. The token “Musterforname_2” is assigned to the class “First name”, the token “Modellername_2” to the class “Last name”, the token “Musterstr._2” to the class “Street” and the token “Musterstadt_2” to the class “City”. The classified tokens 109 of the data set DS2 are each in the form of a class assignment 113 one of the by the trigger definition 123 associated with defined classes and stored with a pointer 115 to their location in the first data model, ie DS2 , connected.

This means that all tokens of the second record are DS2 likewise in a redundancy-free form with their class assignments in the second data model 110 linked to a pointer to their location in the first data model.

13 FIG. 10 shows a schematic block diagram of exemplary data processing by the multi-model database management system and the learning module. There will be a predetermined trigger definition 123 of the form "[Trigger1 = +, Radius = NP] [Trigger1 = x, Radius = NP] [Class]". This trigger definition 123 serves to classify tokens generated from an image file, the image file being broken down into tokens in the form of pixel groups. The trigger definition 123 defines two triggers, ie a first trigger in the form of a pixel group with the content “+” and a second trigger in the form of a pixel group with the content “x”. In addition, the trigger definition defines that a pixel group which is arranged within a first radius of N pixels around the first trigger and at the same time within a second radius of N pixels around the second trigger is a token of the class " Class "acts.

In a document-based data model 106 a database is a document 108 saved. This document 108 comprises a data set DS. For example, the data record DS is a two-dimensional image file. This image file is broken down into tokens by means of a tokenizer, whereby the tokens are pixel groups 150 acts. For example, the pixel groups of equal size in Z by Z are broken down. The tokens include, for example, a first token in the form of a pixel group with the content “x”, a second token in the form of a pixel group with the content “+”, a third token in the form of a pixel group with the content “#” and a fourth token in Shape of a pixel group with the content "-".

The two tokens "+" and x "are used as triggers 121 according to the trigger definition 123 identified. Using the identified triggers and the trigger definition 123 becomes the third token "#" of the class defined by the trigger definition 111 assigned as it is within a first radius in the two-dimensional image file 152 of N pixels around the first trigger “+” and at the same time within a second radius 154 of N pixels around the second trigger "x". Since the fourth token "-" is not under the trigger definition 123 falls, it is assigned to the receiving class as an unknown date.

The triggers 121 Identified tokens “+” and “x” as well as the token “#” classified on the basis of these triggers and the token “-” assigned to the trap class are in an index in a second data model 110 saved. The triggers “+” and “x” are used in the form of a trigger assignment 117 the trigger definition 123 assigned. The token "#" is in the form of a class assignment 113 that of the trigger definition 123 saved assigned to defined classes. The token "-" is in the form of an assignment 119 assigned to the collection classes. In addition, all triggers and classified tokens are in the second data model 110 with a pointer 115 linked to their storage location in the first data model, ie DS.

14th Figure 3 shows a schematic block diagram of one embodiment of an exemplary source computer system 100 as well as an exemplary receiving computer system 200 of a distributed database system 170 . The embodiment from 14th largely correspond to the embodiments from 8th . In 14th assign the two computer systems 100 , 200 additionally one learning module each 120 , 220 which is next to the tokenizer 122 , 222 additional trigger definitions 123 , 223 to classify the data by means of a classifier 124 , 224 as well as a statistical model 125 , 225 .

15th Figure 12 shows a flow diagram of an embodiment of another exemplary method for implementing a cross-database index on a distributed database system. In block 600 For a source database of the distributed database system, a pre-trained learning module for machine learning is provided, which includes a plurality of predetermined trigger definitions. These predetermined trigger definitions define triggers for an assignment of Tokens to classes of a group of classes. In block 602 the corresponding source database is made available. The source database is managed by a multi-model database management system and comprises a plurality of data records which are stored in a document-oriented data model. These saved records each include one or more fields with field values. In addition, the source database provided includes a searchable index of all data comprised by the stored data records. This index is stored redundancy-free in a further data model managed by the multi-model database management system. The index comprises a plurality of tokens generated from the field values of the stored data records which are each linked in the index with one or more pointers to one or more of the data records and / or fields stored in the document-oriented data model, from whose field values the corresponding token was generated .

First tokens in the index, which are included in one of the trigger definitions as triggers, are each assigned to the corresponding trigger definition. Second tokens in the index are each assigned to one or more classes of the group of classes. The remaining tokens in the index are ultimately assigned a trap class to identify the corresponding remaining tokens as unknown data. The assignment to the collecting class excludes an assignment to one of the trigger definitions as well as an assignment to one of the classes of the first group of classes.

In block 604 an additional data record is received and in block 606 stored by the multi-model database management system in the document-oriented data model of the source database. The storage takes place in a document or data container. The additional data set is processed further using a data processing function. This includes: In block 608 one or more additional tokens are generated from additional field values which the additional data record comprises. In block 610 one or more first additional tokens are identified as triggers if they are included as triggers in one of the trigger definitions. In block 612 the remaining additional tokens are classified. The one in block 610 identified triggers are used to assign one or more second additional tokens to one or more classes of the group of classes, if the corresponding second additional tokens from the additional data set in a combination with one or more of the identified triggers according to one of the trigger definitions and the corresponding triggers trigger a corresponding class assignment according to the corresponding trigger definition. The remaining additional tokens, for which no assignment to one of the trigger definitions and no class assignment based on one of the trigger definitions has been made, are in the course of the classification in block 612 assigned to the reception class. Finally, in block 614 the index by the multi-model database management system using the using the additional tokens from block 608 , the class assignments of the additional tokens from block 612 and added a pointer to the additional data record stored in the document-oriented data model. If pointers indicate individual fields of the additional data record, a plurality of pointers is used for a plurality of fields.

You can add in block 614 include matching the additional tokens with the index. If one of the additional tokens is not included in the index, the corresponding additional token is added to its class assignments in the index and linked to the pointer to the additional data record stored in the document-oriented data model. If one of the class assignments of an additional token included in the index is not included in the index, the corresponding class assignment is supplemented with the corresponding additional token in the index and the corresponding additional token is stored in the index with the pointer to the additional in the document-oriented data model Record linked. If one of the additional tokens with all of its class assignments is included in the index, the corresponding additional token in the index is linked to the pointer to the additional data record stored in the document-oriented data model.

You can also complete in block 614 an identification of combinations of second additional tokens with one or more of the identified triggers, which have triggered a class assignment according to one of the trigger definitions, in the index as classified combinations. Class assignments are only carried out for combinations of second additional tokens and one or more identified triggers that are not marked as classified combinations. Thus, by means of the identification, it can be avoided that the classification is repeated for already known and classified combinations with a repeated occurrence in different data records. Rather, before classification, token combinations are compared with the index. If the index already includes the corresponding token combination and this is marked as classified, there is no new classification for this token combination. It only the corresponding token combination and / or the partial combinations and individual tokens comprised by the corresponding token combination are linked in the index with the pointer to the additional data record stored in the document-oriented data model.

In block 616 a supplementary data record resulting from the application of the data processing function is created. This supplementary data record includes the supplements made in the index of the source database and is used to supplement at least one second index of at least one second individual database of the distributed database system as a receive database. In block 618 the supplementary data record is sent via a network to the second individual database for integration into the second index of the receiving database.

16 Figure 12 shows a flow diagram of an embodiment of an exemplary method for generating combined trigger definitions. In block 700 one or more trigger combinations are identified by the learning module, each of which is comprised by at least one of the data records and fulfills a combination criterion. In block 702 will be for each of the in block 700 identified trigger combinations, the trigger definitions of the triggers of the corresponding trigger combinations are combined into one or more additional combined trigger definitions. In block 704 the plurality of predetermined trigger definitions of the learning module is supplemented by the one or more additional combined trigger definitions.

17th Figure 3 shows a flow diagram of an embodiment of an exemplary method for adding additional trigger definitions. In block 800 the pre-trained learning module is supplemented by one or more additional trigger definitions. The additional trigger definitions define additional triggers for a replacement of assignments of tokens in the index to the receiving class by assignments to one or more classes of a further group of classes in the course of a reclassification. The additional trigger definitions can be received by the learning module, for example. For example, the corresponding additional trigger definitions are provided by an administrator. According to alternative embodiments, the additional trigger definitions to be supplemented are created by the learning module. For this purpose, the learning module comprises a statistical model which is used for a statistical analysis of the tokens included in the collection classes and their occurrence in the data records. The result of the statistical analysis is used to create the additional trigger definitions to be added.

In block 802 one or more tokens assigned to the trap class are reclassified in the index, which tokens defined the additional trigger definitions as additional triggers. The reclassification by the learning module includes replacing the assignment to the collection class with an assignment to the corresponding additional trigger definition, which includes the corresponding token as an additional trigger. In block 804 the additional triggers for reclassifying one or more tokens assigned to the capture class in the index to one or more classes of the further group of classes are used by the learning module if the corresponding tokens assigned to the capture class are from one of the data records in a combination with one or more the additional triggers are included and the corresponding additional triggers trigger a corresponding assignment to the one or more classes of the further group of classes in accordance with the corresponding additional trigger definition.

For example, the method for adding additional trigger definitions can be carried out repeatedly following a recursive scheme. The trigger definitions to be added to each recursion level each include additions to a trigger definition of a preceding recursion level so that the recursive additions form tree structures which each include one of the predetermined trigger definition as a root node.

18th Figure 12 shows a flow diagram of one embodiment of an exemplary method for correcting trigger definitions in block. In block 900 a corrected trigger definition is received to replace one of the stored trigger definitions of the learning module. This corrected trigger definition is provided by an administrator, for example. According to alternative embodiments, the corrected trigger definition is created by the learning module using a statistical model. In block 902 the corresponding saved trigger definition is replaced by the corrected trigger definition. In block 904 the tokens classified using the corresponding stored trigger definition are reclassified, the reclassification taking place using the corrected trigger definition.

List of reference symbols

100

Computer system

102

Storage

104

Source database

106

first data model

108

Records

109

Token

110

second data model

111

great

112

index

113

Class assignment

114

processor

115

pointer

116

Program instructions

117

Trigger assignment

118

MM-DBMS

119

Assignment to collection class

120

Learning module

121

Trigger

122

Tokenizer

123

Trigger Definitions

124

Classifier

125

statistical model

126

Communication interface

130

Supplementary data set

150

Pixel group

152

radius

154

radius

160

Payload

162

Access certificate

164

Access certificate

166

Access certificate

168

User certificate

170

distributed database system

172

ID database

174

User certificate

176

User certificate

200

Computer system

202

Storage

204

Receive database

206

first data model

208

Records

209

Token

210

second data model

211

great

212

index

213

Class assignment

214

processor

215

pointer

216

Program instructions

217

Trigger assignment

218

MM-DBMS

219

Assignment to collection class

220

Learning module

221

Trigger

222

Tokenizer

223

Trigger Definitions

224

Classifier

225

statistical model

226

Communication interface

F1, ..., F8

Fields

DS1, ..., DS3

Records

K ₁ , ..., K ₆

node

D ₁ , ..., D ₆

Data

I ₁ , ..., I ₆

index

E (I ₁ ), E (I ₂ )

Supplementary data set

Claims (32)

Computer-implemented method for implementing a cross-database index (112, 222) on a distributed database system (170) which comprises a plurality of independent individual databases (104, 204), wherein the individual databases are communicatively connected to one another via a network (180), the individual databases are each managed by a multi-model database management system (118, 218), the individual databases each comprising a plurality of database-specific data records (108, 208) which are stored in a document-oriented first data model (106, 206) of the respective individual database, the stored data records each comprise one or more field values, the individual field values of the stored data records each being stored in a field (F1, ..., F8), the individual databases also each comprising a searchable first index (112, 212), which in a second data model (110) of the respective en individual database is stored, the index of the respective individual database being a plurality of tokens (109, 209) generated from the field values of the data records stored in the document-oriented data model of the corresponding individual database, the tokens in the index each having one or more pointers (115, 215) to one or more of the in the document-oriented data model of corresponding individual database stored data records is linked from whose field values the corresponding token was generated, the method comprising: • receiving an additional data record (108) by a first individual database as a source database (104) for supplementing data records of the source database, • saving the additional data record , which comprises one or more additional field values, by a first multi-model database management system (118) of the source database in a document-oriented first data model (106) of the source database, • applying a data processing function of the source database to the additional data record, the data processing The processing function comprises at least one indexing of the additional data set for storage in a second data model (110) of the source database, the indexing comprising: daten the source database generating one or more additional tokens (109) from the additional field values, ◯ supplementing the index (112 ) the source database by the first multi-model database management system using the additional tokens and a pointer (115) to the additional data set stored in the document-oriented data model of the source database, • creating a supplementary data set (130) resulting from the application of the data processing function, which the includes additions made in the index of the source database to add to at least one second index of at least one second individual database of the distributed database system as a receive database, sending the supplementary data set over the network to the second individual database for integration into the second index (212) of the receiving database (204). Computer-implemented method according to Claim 1 The supplementing of the index comprises: • Matching the additional tokens with the index of the source database, • If one of the additional tokens is not included in the index of the source database, supplementing the corresponding additional token in the index of the source database and linking the corresponding additional token in the index with the pointer to the additional data set stored in the document-oriented data model of the source database, if one of the additional tokens is included in the index of the source database, linking the corresponding additional token in the index of the source database with the pointer to the additional one in the document-oriented data model of the source database stored data record. Computer-implemented method according to one of the preceding claims, wherein the supplementary data set is sent from the source database via the network to a predefined first group of one or more individual databases of the plurality of individual databases as receive databases which the second individual database comprises. Computer-implemented method according to one of the Claims 1 to 2 , wherein the supplementary data set is sent from the source database via the network to all further individual databases comprised by the plurality of individual databases as receive databases. Computer-implemented method according to one of the preceding claims, wherein a predefined second group of a plurality of individual databases of the plurality of individual databases is configured to function as a source database in each case when additional data records are received. Computer-implemented method according to one of the Claims 1 to 4th , wherein all individual databases of the plurality of individual databases are configured to act as a source database when additional data records are received. Computer-implemented method according to one of the preceding claims, wherein a sensor is assigned to the source database, wherein the additional data set comprises sensor data acquired by the sensor. Computer-implemented method according to one of the preceding claims, wherein each individual database of the plurality of individual databases is assigned one or more sensors, from which the respective individual database receives recorded sensor data in the form of additional data records. Computer-implemented method according to one of the preceding claims, wherein the data processing function further comprises normalizing the additional tokens in the course of adding to the index of the source database. Computer-implemented method according to Claim 9 , the normalization satisfying the fifth and / or sixth normal form. Computer-implemented method according to one of the preceding claims, wherein the data processing function further comprises assigning rights to manage and / or process the data of the additional data set. Computer-implemented method according to Claim 11 , wherein the first multi-model database management system assigns an owner right with respect to the additional data record and / or the supplementary data record to a first entity assigned to the source database. Computer-implemented method according to Claim 12 wherein the first entity is a first user of the source database or a first computer system assigned to the source database. Computer-implemented method according to one of the Claims 11 to 13 , wherein access authorization certificates (162, 164, 166) are stored in the source database as part of the additional data record and / or the supplementary data record, the access authorization certificates comprising one or more of the following access authorization certificates: allows on the content of a record; • a write access credential (164) that enables modifying access to the contents of a data set; An index access authorization certificate (166), which enables knowledge of the existence of the data record in the database containing the data record and read access to metadata of the data record. Computer-implemented method according to Claim 14 wherein a read access credential (162), a write access credential (164) and an index access credential (166) are stored as part of the additional data set and / or the supplementary data set for the first entity. Computer-implemented method according to one of the Claims 11 to 15th wherein a second entity associated with the receive database is further associated with an index access credential (166) for the additional record. Computer-implemented method according to Claim 16 , wherein the second entity is a second user of the receive database or a second computer system assigned to the receive data. Computer-implemented method according to one of the Claims 16 to 17th wherein a prerequisite for sending the supplementary data record to the receiving database is a successful check that the second entity is assigned an index access authorization for the supplementary data record. The computer-implemented method according to one of the preceding claims, wherein the data processing function further comprises classifying the tokens generated from the additional data set. Computer-implemented method according to Claim 19 , wherein the source database for the classification comprises a pre-trained learning module (120) for machine learning, wherein the pre-trained learning module comprises a plurality of predetermined trigger definitions (123) which triggers (121) for assigning tokens to classes of a first group of Define classes, with first tokens in the index of the source database, which are included in one of the trigger definitions of the source database as a trigger, are each assigned to the corresponding trigger definition, with second tokens in the index of the source database each one or more classes of the first Group of classes are assigned, wherein the classifying comprises: • if one or more first additional tokens are included as a trigger by one of the trigger definitions, the learning module identifies the corresponding token as a trigger, • using the identified triggers for assigning one or more several second additional tokens to one or more Kla ssen the first group of classes by the learning module if the corresponding second additional tokens are included in the additional data record in a combination with one or more of the identified triggers according to one of the trigger definitions, the corresponding triggers according to the corresponding trigger definition a trigger corresponding class assignment, the addition of the index by the first multi-model database management system using the class assignments of the additional tokens. Computer-implemented method according to Claim 20 , with remaining tokens that are not assigned to either the trigger definition or one or more classes of the first group of classes in the index of the source database for identifying the corresponding remaining tokens as unknown data being assigned to a collection class, the assignment to the collection class being a Assignment to one of the trigger definitions as well as an assignment to one of the classes of the excludes first group of classes, wherein the classification further comprises: Assigning the remaining additional tokens, for which no assignment to one of the trigger definitions and no class assignment has been made based on one of the trigger definitions, to the collection class. Computer-implemented method according to one of the Claims 20 to 21st wherein, if one of the additional tokens is not included in the index, the corresponding additional token is supplemented under its class assignments in the index, and if one of the class assignments of an additional token included in the index is not included in the index Supplementing the corresponding class assignment with the corresponding additional token in the index and linking the corresponding additional token in the index with the pointer to the additional data record stored in the document-oriented data model takes place. The computer-implemented method according to any one of the preceding claims, wherein the pointers with which the tokens are stored linked in the index each refer to one or more of the field values in the stored data records. Computer-implemented method according to one of the preceding claims, wherein the field values of the additional data record include text data, image data, audio data and / or video data. A computer-implemented method according to any one of the preceding claims, wherein generating the tokens comprises applying tokenization logic (122) to the field values of the additional data set which comprises a full-text indexer configured to break text into words and to output the words as tokens. Computer-implemented method according to one of the Claims 1 to 24 wherein generating the tokens comprises applying tokenization logic (122) to the field values of the additional data record, which logic comprises a generic tokenizer which is configured to recognize data of different data types in the field values and to generate from these tokens in different data types. Computer-implemented method according to one of the preceding claims, wherein at least the document-based data model used by the multi-model database management system for storing the data sets is a NoSQL data model. Computer-implemented method according to one of the preceding claims, wherein the machine learning is configured for data extraction, consistency checking, image recognition, speech recognition, voice control, device monitoring and / or autonomous device control. The computer-implemented method of any of the preceding claims, the method further comprising: Receipt of a further supplementary data record generated by a third individual database as an additional source database by the first individual database via the network, the further supplementary data record comprising additions made in an index of the additional source database, • Integrating the further supplementary data set into the index of the first individual database, the integration comprising supplementing the corresponding index in which ◯ a comparison of tokens of the further supplementary data set with the index of the source database, ◯ if one of the tokens of the further supplementary data set is not included in the index of the source database, supplementing the corresponding token in the index of the source database and linking the corresponding token in the index with a pointer to the data set stored in a document-oriented data model of the additional source database to which the corresponding token was generated, ◯ if one of the tokens of the additional supplementary data set is included in the index of the source database, linking the corresponding token in the index of the source database with the pointer to the data set stored in a document-oriented data model of the additional source database from which the corresponding token was generated. The computer-implemented method of any of the preceding claims, the method further comprising: • Receipt of a search query by the first individual database, the search query containing one or more search values, • Searching the index of the first individual database for the search value, • Identify one or more tokens within the index that match one or more of the search values, Returning a response to the search query, the response comprising at least: an indication of the identified tokens, one or more specific data records or one or more references to the specific data records obtained by analyzing pointers with which the identified tokens are associated. Source computer system (100) for implementing a cross-database index (112, 222) on a distributed database system (170) which comprises a plurality of computer systems (100, 200) comprising the source computer system, each with an independent individual database (104, 204), the computer systems are communicatively connected to the individual databases via a network (180), the computer systems also having one or more processors (114, 214), one or more data storage media (102, 202) on which the respective individual database is provided, a communication interface ( 126, 226) for communication via the network and a program logic (116, 216), the individual databases each being managed by a multi-model database management system (118, 218), the individual databases each having a plurality of database-specific data records (108, 208 ), which in a document-oriented first date n model (106, 206) of the respective individual database are stored, the stored data records each comprising one or more field values, the individual field values of the stored data records each being stored in a field (F1, ..., F8), the individual databases also each comprise a searchable first index (112, 212) which is stored in a second data model (110) of the respective individual database, the index of the respective individual database being a plurality of the field values of the document-oriented data model of the corresponding individual database includes generated tokens (109, 209), the tokens in the index being linked to one or more pointers (115, 215) to one or more of the data records stored in the document-oriented data model of the corresponding individual database, the corresponding token was generated from their field values, the program logic (116) of the source computer system being configured to execute a method for implementing a cross-database index, the method comprising: receiving an additional data record (108) by a first individual database as the source database (104 ) for supplementing data records of the source database, • storing the additional data record, which includes one or more additional field values, by a first multi-model database management system (118) of the source database in a document-oriented first data model (106) of the source l database, • applying a data processing function of the source database to the additional data set, wherein the data processing function comprises at least one indexing of the additional data set for storage in a second data model (110) of the source database, wherein the indexing comprises: ◯ generating one or more additional tokens (109 ) from the additional field values by the source database, ◯ supplementing the index (112) of the source database by the first multi-model database management system using the additional tokens and a pointer (115) to the additional data set stored in the document-oriented data model of the source database, Creation of a supplementary data record (130) resulting from the use of the data processing function, which comprises the supplements made in the index of the source database, for supplementing at least one second index of at least one second individual database of the distributed database system a As a receiving database, sending the supplementary data set via the network to the second individual database for integration into the second index (212) of the receiving database (204). Distributed database system (170) for implementing a cross-database index (112, 222) on the distributed database system, which comprises one of the plurality of computer systems, each with an independent individual database, the computer systems with the individual databases being communicatively connected to one another via a network, wherein the computer systems furthermore each have one or more processors (114, 214), one or more data storage media (102, 202) on which the respective individual database is provided, a communication interface (126, 226) for communication via the network and a program logic (116 , 216) include, the individual databases each being managed by a multi-model database management system (118, 218), the individual databases each comprising a plurality of database-specific data records (108, 208) which are stored in a document-oriented first data model (106, 206) of the respective individual database , the stored data records each comprising one or more field values, the individual field values of the stored data records each being stored in a field (F1, ..., F8), wherein the individual databases each further comprise a searchable first index (112, 212) which is stored in a second data model (110) of the respective individual database, the index of the respective individual database being a plurality of the field values in the document-oriented data model of the corresponding individual database stored data records includes generated tokens (109, 209), the tokens in the index each being linked to one or more pointers (115, 215) to one or more of the data records stored in the document-oriented data model of the corresponding individual database, from whose field values the corresponding Token has been generated, wherein the program logic is each configured to execute a method for implementing a cross-database index by the computer system executing the program logic of the plurality of computer systems, which acts as a source computer system (100), the method comprising t: • Receipt of an additional data record (108) by a first individual database as a source database (104) for adding data records to the source database, • Saving the additional data record, which comprises one or more additional field values, by a first multi-model database management system (118) of the source database in a document-oriented first data model (106) of the source database, • Applying a data processing function of the source database to the additional data record, the data processing function comprising at least one indexing of the additional data record for storage in a second data model (110) of the source database, the indexing comprising: ◯ Generation of one or more additional tokens (109) from the additional field values by the source database, ◯ Supplementing the index (112) of the source database by the first multi-model database management system using the additional tokens and a pointer (115) to the additional data set stored in the document-oriented data model of the source database, • Creation of a supplementary data record (130) resulting from the use of the data processing function, which comprises the additions made in the index of the source database, for supplementing at least one second index of at least one second individual database of the distributed database system as a receive database, • Sending the supplementary data set via the network to the second individual database for integration in the second index (212) of the receiving database (204).

Images