WO2013095084A1 - Ontology-based enterprise application integration - Google Patents

Abstract

An Enterprise Application Integration system (102) comprising a workflow module (106), and at least one connector (108) to one or more further modules, wherein data is mapped between different modules (14, 18, 22, 26) by reference to semantic data mapping rules defined by an ontology (110)

Description

ONTOLOGY-BASED ENTERPRISE APPLICATION INTEGRATION

Field of Invention

The invention relates to an Enterprise Application Integration (EAI) system.

Background

Enterprise Application Integration is a way of linking systems and applications in an organisation, so the data can be passed therebetween. The systems and applications may reside on different operating systems, use different database solutions and different computer languages, and in some cases are legacy systems that are no longer supported by the vendor who originally created them. Therefore the EAI forms a communication interface between them, sometimes referred to as middleware, to enable integration of the systems and applications across the organisation.

A conventional EAI will make use of connectors in the platform for integrating with different application databases/modules. For example, a MsSQL connector may be required to integrate the Enterprise Resource Planning (ERP) database, an SAP connector for the Customer Relationship Management (CRM) database, a DB2 connector for the Finance database, and an Oracle connector for the Business Intelligence database.

However, there are a number of problems with this approach. For example, if there is an update to the architecture of one of the databases, the corresponding connector will need to be recoded to maintain access. When the recoded connector is installed into the EAI there are sometimes unforeseen knock-on effects which can be detrimental and affect other parts of the system. Similarly, if there is a requirement to connect to a new database, a new connector has to be coded, which often involves a steep learning curve, and has potentially similar unwanted effects. Creating or updating a connector is time-consuming, as the programmer typically has to manually match the data fields in the different databases or otherwise ensure they are correctly interpreted, which of course can be quite difficult as the database architectures are usually all different.

Thus an aim of the invention is to reduce the risk of unwanted side-effects on an EAI system when an application is updated.

A further aim of the invention is to reduce the amount of time taken to integrate different applications with an EAI.

Summary of Invention In an aspect of the invention, there is provided an Enterprise Application Integration system comprising a workflow module, and at least one connector to one or more further modules, characterised in that data is mapped between different modules by reference to semantic data mapping rules defined by an ontology.

Typically the further modules are external databases and/or applications.

As data is mapped automatically between the different modules by reference to the rules for interpreting the data, it is not necessary to recode a connector if the corresponding module structure changes, as the system automatically adapts in response to such a change.

In one embodiment the connector is a web service connector for connecting to the web service of each further module. Typically the web service connector defines each further module via a Uniform Resource Locator (URL).

As a result, if a module structure is updated, only the URL in the EAI requires updating, as the code language of the web service is standardised as Extensible Markup Language (XML). In other words there is no need to interpret the internal language of the module or application being connected to. Advantageously there are no side-effects as the structure of the EAI is not altered by importing new code, as would be required in a conventional system. In one embodiment the web service is identified from the ontology. Typically the ontology includes a plurality of URLs. Thus the EAI refers to the ontology to identify where to get the relevant data from. In one embodiment the ontology uses a Web Ontology Language (OWL). Typically the data mapping rules include inferring identity between data of different modules using any or any combination of the axioms of owl: same As, owhFunctionalProperty and/ or owl : InverseFunctionalProperty . In one embodiment the data mapping rules include the use of words from at least one linguistic knowledge base.

In one embodiment one or more processes are extracted from the workflow module and for each process the web service of a corresponding further module is identified.

In a further aspect of the invention, there is provided a method of mapping data in an Enterprise Application Integration system comprising the steps of:

extracting one or more processes from a workflow module;

using at least one connector to obtain data for the processes from one or more further modules;

characterised by mapping data between different modules by reference to semantic data mapping rules defined by an ontology. In one embodiment the web service of a corresponding further module is identified for providing data to a specified process.

In one embodiment a web service connector is used to obtain data from the web service of a further module.

In one embodiment if the structure of a further module is changed, the corresponding web service connector can be updated by changing the URL.

In a yet further aspect of the invention there is provided an Enterprise Application Integration system characterised in that a web service connector is provided for connecting to the web service of one or more further modules, each of which is defined via a Uniform Resource Locator.

Brief Description of Drawings

It will be convenient to further describe the present invention with respect to the accompanying drawings that illustrate possible arrangements of the invention. Other arrangements of the invention are possible, and consequently the particularity of the accompanying drawings is not to be understood as superseding the generality of the preceding description of the invention.

Figure 1 is a schematic view of a conventional Enterprise Application Integration system according to the prior art. Figure 2 is a schematic view of a Enterprise Application Integration system according to an embodiment of the invention. Figure 3 is a schematic view of a method of mapping data in an Enterprise Application Integration system according to an embodiment of the invention.

Detailed Description With reference to the prior art system illustrated in Figure 1, a conventional EAI 2 comprises a server 4 which is connected to a workflow engine module 6 for defining the workflow processes, and a data mapping file 10. The EAI may also include a communications server and a business rules engine (not shown). The server 4 includes a number of connectors to different data sources, namely a MsSQL connector 12 for the ERP module 14, an SAP connector 16 for the CRM module 18, a DB2 connector 20 for the Finance module 22, and an Oracle connector 24 for the Business Intelligence module 26. The connectors allow a direct connection to be made to obtain data from the data sources in the relevant language. The data mapping file 10 includes data mappings which are manually created to map the data from the data sources to the workflow engine module. However, if the structure of a data source changes, the corresponding connector has to be rewritten, and the data mapping file has to be manually updated, both of which can be difficult and time-consuming. With reference to Figure 2, the invention is illustrated wherein the EAI 102 comprises a server 104 connected to a workflow engine module 106 for defining the workflow processes substantially as before.

However, instead of multiple application-specific independent connectors, the EAI 102 includes a web service connector 108 for connecting to the web service of each of the data sources 14, 18, 22, 24. The web service connector simply specifies the URL for each of the data sources, such that if there is a change of a data source changes, only the URL requires updating. In addition, the data is mapped between different modules by reference to semantic data mapping rules defined by an ontology 110. The EAI can also include a connection to a third party Decision Support System 28 for enhancing the accuracy of the data mapping rules. Thus the data is mapped automatically, rather than manually.

Figure 3 provides further detail on how data is mapped in an EAI. Application data is automatically mapped 230, then the workflow is extracted 242 from the EAI ontology from inferred information wherein the extracted information may have new relationships reasoned and created at runtime. Next, the defined processes are executed 252 and the web services are called 254, then advice is provided 256 based on analysed data by triggering the Decision Support web service, and finally an Email or SMS can be sent 270 whenever required in the workflow.

The automatic mapping process 230 comprises the following steps:

• Retrieve instances 232 that are inferred to be the same from ow sameAs e.g.

Finance:accNo is considered to be the same as Procurement:accountID axiom;

• Retrieve instances 234 that are inferred to be the same from owl:FunctionalProperty e.g. Finance :hasTotalBalance is considered to be a type of owl:FunctionalProperty because there can only be one relationship axiom;

• Retrieve instances 236 that are inferred to be the same from owl:InverseFunctionalProperty axiom e.g. Procuremen hasPurchaseOrder is a type of owhlnverseFunctionalProperty because a person can have multiple purchase orders, but each purchase order can only belong to one individual;

• Expand instances 238 to include words from linguistic knowledge base (e.g.

Wordnet);

• Identify all instances considered to be the same and relate them to the web service parameters to be triggered 240. Thus semantics are used to correlate different fields such as 'Customer No.' and 'Customer ID' due to the rules provided in the ontology. Thus the fields from the databases and/or workflow can be mapped and those required are retrieved. The workflow extraction process 242 comprises the following steps to identify the corresponding fields in the workflow and match them to those in the databases:

• Extract processes 244 from workflow ontology;

• Identify suitable web service 246 for the process based on criteria defined as rules;

· Extract input parameters 248 (and related instances output from the fields mapper);

• Map output parameters 250 to input parameters of the next process as defined in the ontology. After the data is matter, a decision support process is activated. The advice process 256 comprises the following steps:

• Extract required relevant information 258 from database based on the input key;

• Apply business rules 260 on the extracted data for data transformation and filtration;

• Apply the transformed data 262 to decision model which can be based on mathematical model or logic-based;

• Execute the decision model 264; • Apply business rules 266 on the results returned by the previous step;

• Return the filtered results 268 to the workflow.

It will be appreciated by persons skilled in the art that the present invention may also include further additional modifications made to the device which does not affect the overall functioning of the device.

Claims

Claims

1. An Enterprise Application Integration system (102) comprising a workflow module (106), and at least one connector (108) to one or more further modules, characterised in that data is mapped between different modules (14, 18, 22, 26) by reference to semantic data mapping rules defined by an ontology (110).

2. A system (102) according to claim 1 wherein a web service connector (108) is provided for connecting to the web service of each further module (14, 18, 22, 26).

3. A system (102) according to claim 2 wherein the web service connector (108) defines each further module (14, 18, 22, 26) via a Uniform Resource Locator. 4. A system (102) according to claim 2 or 3 wherein the web service is identified from a plurality of Uniform Resource Locators listed in the ontology (110).

5. A system (102) according to any preceding claim wherein the data mapping rules include inferring identity between data of different modules (14, 18, 22, 26) using any or any combination of a linguistic knowledge base and/or the

Web Ontology Language axioms of owl: same As, owLFunctionalProperty and/or owl : InverseFunctionalProperty . A system (102) according to any preceding claim wherein one or more processes are extracted from the workflow module (106) and for each process the web service of a corresponding further module (14, 18, 22, 26) is identified.

A system (102) according to any preceding claim wherein the further modules are external databases and/or applications.

A method of mapping data in an Enterprise Application Integration system (102) comprising the steps of:

extracting one or more processes from a workflow module (106);

using at least one connector (108) to obtain data for the processes from one or more further modules (14, 18, 22, 26);

characterised by mapping data between different modules (14, 18, 22, 26) by reference to semantic data mapping rules defined by an ontology (110).

A method according to claim 8 wherein the web service of a corresponding further module (14, 18, 22, 26) is identified for providing data to a specified process.

10. A method according to claim 9 wherein a web service connector (108) is used to obtain data from the web service of a further module (14, 18, 22, 26).

11. A method according to claim 10 wherein if the structure of a further module (14, 18, 22, 26) is changed, the corresponding web service connector can be updated by changing the Uniform Resource Locator. 12. An Enterprise Application Integration system (102) characterised in that a web service connector (108) is provided for connecting to the web service of one or more further modules (14, 18, 22, 26), each of which is defined via a Uniform Resource Locator.