KR20050021571A - Database model for hierarchical data formats - Google Patents

Abstract

The present invention relates to a method of mapping a hierarchical data format to a relational database management system. It is an object of the present invention to insert descriptors 1, 10, 11, read portions of descriptors 1, 10, 11, read entire descriptors 1, 10, 11, and perform a quick text query. To provide a method for mapping hierarchical data formats including descriptors (1, 10, 11) to a relational database management system that can handle various types of hierarchical descriptors (1, 10, 11) in a fast manner. will be. According to the invention, the descriptors 1, 10, 11 are separated into parts of a common format, which are stored in the relationships 20, 21, 22... In the relational database.

Description

DATABASE MODEL FOR HIERARCHICAL DATA FORMATS}

The present invention relates to a method of mapping a hierarchical data format to a relational database management system. Moreover, the present invention relates to a database model and apparatus for reading from and / or writing to a recording medium using such a method.

The future of digital recording is characterized by the preparation, presentation and attainment of value-added data services, i.e. recorders such as digital video recorders (DVRs) are provided by content providers or special services such as broadcast stations. It will store and process additional information passed or even assembled by the user himself. Value added (metadata) is created to provide another user with information. For example, the added value may be a movie summary describing the plot, a list of actors, and the like. In addition, the provision of additional information to facilitate navigation within the movie constitutes added value. For example, a movie can be made up of sections, subsections, and the like, each section and subsection having a separate title and possibly containing another useful information.

Hierarchical data formats are generally used to provide structural information and to transmit other metadata for multimedia objects such as video or audio streams. A well known and widely accepted hierarchical data format is XML (eXtensible Markup Languages). XML is a system that restricts the special markup language used to send formatted data. Therefore, XML is called a meta language and is the language used to create other special languages. XML data consists of text organized in the form of a plurality of descriptors. The text itself contains elements, attributes and content, i.e. the remaining text. In addition to the use of multimedia objects, many other applications for XML are known.

In the foreseeable future, digital recorders store very large amounts of data or other hierarchical data formats in XML in relational databases because it is widely used and very complex. However, a problem arises that the hierarchical data format must be mapped to a relational database management system (RDBMS) for storage. Many database models for XML have already been proposed. For example, Rahaayu et al., Representation of multilevel composite objects in relational database (1998 International Conference on Object-Oriented Information Systems, Bulletin of OOIS'98, pp 221-238), or Zhang. See, On Supporting Containment Queries in Relational Database Management Systems (2001, ACM.Sigmod Record, Vol. 30, No. 2, pp. 425-36). However, no database model is known that can handle various types of hierarchical descriptors as a quick way to insert descriptors, read portions of descriptors, read entire descriptors, and perform quick text queries.

1 a) and 1 b) are schematic XML descriptors and representations thereof as XML trees;

2 illustrates a database model according to the invention using a single relationship.

FIG. 3 is a database model as in FIG. 2, but showing a database model in which additional information is stored on the descriptor structure. FIG.

4A and 4B are diagrams showing an XML descriptor as in FIG. 1, but showing an XML descriptor in which the text includes a string value and an integer value.

FIG. 5 is a database model similar to FIG. 2 but illustrating a database model in which elements, attributes, integer values, and string values are separated into different relationships. FIG.

FIG. 6 is a database model similar to FIG. 3, but showing a database model in which iterations within a relationship are removed by providing additional relationships.

FIG. 7 illustrates a generic metadata descriptor including namespace information, a unique identifier, and a link to another metadata descriptor.

8 illustrates a metadata stream comprising a plurality of metadata descriptors.

9 shows a database model according to FIG. 6 including a descriptor index;

It is therefore an object of the present invention to provide a method of mapping a hierarchical data format comprising a descriptor to a relational database management system. It is another object of the present invention to provide a database model and apparatus for reading from and / or writing to a recording medium using such a method.

In accordance with the present invention, descriptors are separated into parts of a common format, which parts of the common format are stored in a relationship in a relational database. The method has the advantage of being independent of the structure of the stored descriptor. Only a limited number of common formats are needed to store all types of descriptor formats. Common formats include, for example, elements, attributes, text, and the like. In this way, each descriptor is analyzed word by word, separated into different components, and stored in a relationship that is preferably a table.

The method can be further improved by providing an independent relationship to the common format. All queries use only these relationships. For example, the first relationship includes text only, while the second relationship includes elements and the like. This makes the query quick and simple due to the limited number of relationships. For example, if a text query is to be performed, only relationships that contain text should be retrieved. It is advantageous to provide independent relationships for all common formats, but it is likewise possible to use relationships for more than one common format. For example, elements and attributes may be stored together in a first relationship, while text is stored in a second relationship.

According to a variant of the invention, the method further comprises storing information in the relationship allowing the restoration of the descriptor structure. When a query passes only a single database entry, the complete structure of the descriptor belonging to a particular database entry can be restored.

Advantageously, the information allowing restoration of the descriptor structure includes the descriptor number and relative and / or absolute position of the portion of the common format within the descriptor. Using this information, it is possible to collect appropriate values from the database and classify these values in a useful way. Each time a descriptor is stored in the database, it receives a unique descriptor number. Moreover, for all parts of the descriptor's common format, relative positions in the descriptor and / or absolute positions in the relationship are derived. Descriptor numbers and relative and / or absolute positions are stored in a relationship with parts of the common format.

Advantageously, the information allowing recovery of the descriptor structure further includes an indicator for the next upper hierarchical level of the portion of the common format within the descriptor. This facilitates quick reconstruction of the descriptor part by starting from any part of the descriptor towards the head of the descriptor (level directed). The next upper hierarchical level is information that helps to reconstruct the descriptor portion when only the state or absolute word position of a portion of the common format is known, for example as a query result.

According to another aspect of the invention, the method further comprises storing the descriptor index in a relational database. Such descriptor indexes store additional information about all descriptors and make it easy to find a specific descriptor in the database.

Advantageously, the descriptor index includes at least the descriptor number, the descriptor's absolute position in the relationship and / or unique identifier for the descriptor. Storing this information in the descriptor index allows quick access to specific descriptors in the relationship. The absolute position of the descriptor in the relationship is advantageously defined as the absolute position of the first part of the common format. Because unique identifiers are often needed, faster access to this kind of data is provided by storing the unique identifier in the descriptor index. In addition to the information mentioned, other types of information may be stored in the descriptor index, such as, for example, the descriptor's level number or other useful data.

Advantageously, the hierarchical data format containing the descriptors corresponds to XML. Because XML is widely used and widely accepted, this allows a wide range of applications of the method of the present invention.

According to the invention, the common format includes at least elements, attributes and text. This type of common format is sufficient for many applications. Although elements are primarily used to construct descriptors, text usually contains the information retrieved from the query. Attributes are most often used to characterize an element.

Advantageously, the common format text is further divided into string values and integer values. In this way, a faster search can be achieved because the relationship that must be searched for a query is smaller. For example, a query on a string value is performed in a relationship that contains only string values, which contains fewer elements than a relationship that contains both strings and integer values.

Advantageously, the common format further includes a namespace. This feature is of particular interest to XML and allows the markup intended for one document to avoid conflicts between different documents when using the same element type or attribute name as another document for different purposes.

Advantageously, a database model for mapping a hierarchical data format comprising descriptors to a relational database management system utilizes the method according to the invention. Such a database model can perform simple and fast queries, flexible processing of various descriptor formats, simple and quick reconfiguration of descriptors, and simple and quick insertion of descriptors. Moreover, such a database model can be easily implemented through existing relational database management systems.

Advantageously, an apparatus for reading from and / or writing to a recording medium uses the method or database model according to the invention for mapping a hierarchical data format comprising a descriptor to a relational database management system. Such a device makes it possible to store value added information in an existing relational database. The user of the device can easily use and / or edit the value added information.

In order to better understand the present invention, exemplary embodiments are defined in the following description of advantageous embodiments with reference to the drawings using XML as an example of a hierarchical data format. It is to be understood that the invention is not limited to this exemplary embodiment, and that defined features may also be conveniently combined and / or modified without departing from the scope of the invention.

FIG. 1A shows a schematic example of an XML descriptor 10 and FIG. 1B shows a corresponding representation as an XML tree. As can be seen in the figure, the exemplary embodiment 10 includes sections, subsections and sub-subsections, each having a title. The title of the sub-subsection has an attribute "arrow" with a value "down". The descriptor 10 consists of a total of 17 words, where the text of each title is counted as a single word independently of the actual number of words. For example, "Leonardo is swimming" includes three "real" words, but is a single "logical" word. The number given to each line of the descriptor 10 in a) of FIG. 1 is the relation word position of the first word of each line in the descriptor 10. It can be seen from the corresponding tree structure in FIG. 1 b that the descriptor 10 has five levels, that is, a level from level 0 to level 4. The tree structure is a tool to help illustrate the hierarchical relationship between different words of the descriptor 10.

In figure 2 a database model according to the invention is shown, where a single relationship 20 is used. Relationship 20 is represented by a table. The first column ("value") represents the stored part (XML string) itself. The second column ("Descr #") represents a single descriptor number in the database management system. The column ("word position") contains the relative position of the stored part within the specific descriptor 10. The "Descr #" and "word position" taken together are the primary key of the relationship 20, which allows complete recovery of the descriptor 10. The type of each XML string is stored in the relationship in the column ("type"). In the example, the type includes "element", "attribute" and "text". The last column ("level") contains the hierarchical levels of each XML string as shown in b) of FIG. As can be seen, not all words of the descriptor 10 are stored in the relationship 20. The words "closing" such as </ title> and </ section> do not contain additional information and are not necessary for the recovery of the descriptor 10. Therefore, the "end" word is not stored in the database. Of course, it is also possible to store these words if necessary.

FIG. 3 shows a database model similar to that of FIG. 2, where additional columns (“next top word position”) are included in relationship 21, which relationship 21 is the next top of the XML string in the specific descriptor 10. Contains indicators for hierarchical words. This is information to help recover the descriptor part, for example when only the word position of a part of the common format is known as a query result. Quick reconfiguration of the descriptor portion is facilitated by providing this additional information.

In FIG. 4, another schematic descriptor 11 similar to FIG. 1 is shown. However, in this example, the text consists of a string value and an integer value. As can be seen in FIG. 4 b), the string value and the integer value are separated and counted as separate " logical " words.

FIG. 5 shows a database model similar to that shown in FIG. 2. However, in this example, the XML string is separated into elements, attributes, string values, and integer values, and stored in different relationships 22, 23, 24, 25. This allows for a faster search within the relationships 22, 23, 24, 25. Due to the descriptor number and word position, it is still possible to recover the complete descriptor 11 from different relationships 22, 23, 24, 25. The value ("type") is not necessary for this embodiment, because all relationships 22, 23, 24, 25 contain only a specific type.

In figure 6 further modifications of the database model according to the invention are shown. The database model is similar to that shown in FIG. 3, but the repetition within the relationship 31 is eliminated. This is accomplished by providing additional relationships 32, 33, 34, 35 ("secondary relationships") for elements, string and integer values, and attributes. For each XML string, the value ("type") and the corresponding descriptor key ("Descr.key") are included in the "primary" relationship 31. The descriptor key represents the corresponding entry in the additional relationships 32, 33, 34, 35 for the particular type of XML string. The columns taken together ("type" and "Descr.key") can be considered as secondary keys because the columns associate each XML string defined by the primary key with a particular value.

7 shows a general metadata descriptor 1. The actual content of the metadata descriptor is included in the core 6. Moreover, the metadata descriptor 1 includes a namespace declaration 2, a unique identifier 4, and a link 5 to other metadata descriptors. Namespace declarations 2 and unique identifiers 4 are often stored at specific locations within the database management system because they are needed. The intention is to provide quick access to this kind of data. The namespace declaration (2) is only valid for the specific metadata descriptor (1). The unique identifier 4 allows for explicit identification of the metadata descriptor 1.

FIG. 8 shows a metadata stream 7 comprising a plurality of metadata descriptors 1 as shown in FIG. 7. Moreover, the metadata stream 7 includes a namespace declaration 2, which is valid for all metadata descriptors 1 within a particular metadata stream 7.

In FIG. 9, the use of the descriptor index 40 is shown. The descriptor index 40, for each descriptor stored in the database, includes a descriptor number, the number of descriptor levels ("Max Level"), a unique identifier ("UUID"), and an absolute position in the relationship 41 ("Abs.Pos"). ). The corresponding relationship 41 is similar to that shown in FIG. 6. However, the relationship 41 further includes the absolute location and namespace declaration of each XML string. Additional relationships, including elements, string values, integer values, etc., addressed by secondary keys, are not shown for simplicity.

The database model shown in the figures has a number of advantages:

Flexibility to store all kinds of descriptors by providing separation of the incoming XML stream into a common format.

Fast queries due to a limited number of relationships. For example, a text query should only be performed in a small number of relationships, such as "string values" or "elements", ie only in those relationships in which strings are stored.

Fast implementation of such a database model within a database management system due to a limited number of relationships. Different database models require at least one relationship for each descriptor type.

Fast restoration of descriptors back to XML format due to the specific modeling of the database, ie by using attributes ("Descr #" and "word position").

-Fast recovery of the descriptor part by providing additional information ("next upper word position"). This is helpful when starting from any part of the descriptor towards the head of the descriptor (level oriented).

As noted above, the present invention relates to a method of mapping a hierarchical data format to a relational database management system, and furthermore, the present invention relates to a database model for reading from and / or writing to a recording medium using such a method. And devices.

Claims (13)

A method of mapping a hierarchical data format including descriptors (1, 10, 11) to a relational database management system, Separating the descriptors (1, 10, 11) into parts of a common format, Storing portions of the common format as relations (20, 21, 22 ...) in the relational database And a hierarchical data format comprising a descriptor to a relational database management system. 2. The relational data format of claim 1, further comprising providing an independent relationship (22, 23, ..., 32, 33, ...) to said common format. How to map to a database management system. 3. A hierarchical data format according to claim 1 or 2, further comprising the step of storing information in the relationship (20, 21, 22, ...) that permits recovery of the descriptor structure. How to map to a relational database management system. 4. A method according to claim 3, characterized in that the information allowing restoration of the descriptor structure comprises relative and / or absolute positions of parts of a common format within the descriptors (1, 10, 11), and descriptor numbers, A method of mapping a hierarchical data format containing descriptors to a relational database management system. 5. The information as claimed in claim 4, wherein the information allowing restoration of the descriptor structure further comprises an indicator for the next upper hierarchical level of the portion of the common format in the descriptor (1, 10, 11). A method of mapping a hierarchical data format comprising a descriptor to a relational database management system. 6. A method as claimed in claim 4 or 5, further comprising storing a descriptor index (40) in the relational database. 7. The descriptor index (40) of claim 6, wherein the descriptor index (40) is at least a descriptor number, an absolute position of the descriptor (1, 10, 11) in the relationship (20, 21, 22 ...), and / or the descriptor (1, A hierarchical data format comprising a descriptor, characterized in that it comprises a unique identifier (4) of 10, 11). 8. A hierarchical data format according to any one of the preceding claims, characterized in that the hierarchical data format comprising the descriptors (1, 10, 11) corresponds to XML (eXtensible Markup Language). How to map a hierarchical data format including descriptors to a relational database management system. 9. A method as claimed in any preceding claim, wherein said common format includes at least elements, attributes and text. 10. The method of claim 9, wherein the common format text is divided into a string value and an integer value. 11. A method according to claim 9 or 10, wherein the common format further comprises namespace information (2). A database model for mapping hierarchical data formats including descriptors (1, 10, 11) to a relational database management system, A database model for mapping a hierarchical data format comprising a descriptor to a relational database management system, using the method according to any one of claims 1 to 11. An apparatus for reading from and / or writing to a recording medium, the apparatus comprising: The method of any one of claims 1 to 11, or the database model of claim 12, for mapping a hierarchical data format comprising descriptors (1, 10, 11) to a relational database management system. An apparatus for reading from and / or writing to a recording medium.