OMG Insurance Working Group
An approach to P&C Information Models Development
(Draft)
By
OMG Insurance WG Submission Team
April 22nd2008
Table of Contents
1.Introduction
2.Approach to developing the P&C Information Models
2.1.Available artifacts (models, glossaries…)
2.2.Available tools
2.3.Roles and Responsibilities
2.4.Timeline
2.5.P&C deliverables
2.5.1.P&C Business Glossary
2.5.2.P&C Information models
2.5.3.P&C Ontology
2.5.4.Traceability map
2.6.P&C business function to process to data mapping
1.Introduction
This document provides a high level approach to developing the P&C Information models in response to the request for proposal issued by Object Management Group (OMG, The business glossary from IBM and other sources (as available) will be categorized into a set of subject areas by P&C SMEs working with data modelers experienced in Insurance domain. A base version of the glossary will be maintained as well as additional metadata to better manage and enhance its business value.Working groups that will consist of P&C SMEs and data modelers will work on subject areas (Claim, Underwriting for example) and manage all models in a metadata repository from Adaptive Inc. The proposed P&C business glossary will strive to provide a rich set of P&C business Terms, additional metadata such as source, author, version/history, synonyms, business/geographic context specific variations of the P&C business Terms and pertinent business rules. Such a rich business glossary will promote consistent understanding of P&C business concepts and also demonstrate how OMG standards based tools can help interoperability of P&C metadata across various tools.
A ‘reference subject area/core model’ will be developed (Party, Party Role,Location, Contact Mechanism for example) that can also be used in other domains like Life and Reinsurance. A Traceability map from business glossary to information models will be provided.In addition, XMI representation of information models will also be provided. For additional details on features required for business glossary and information models, please refer to the OMG P&C RFP (link above).
Please refer to Figure 1 for an overview of the proposed P&C standard development process.Please also note that the metadata repository and data modeling tools provided are to be used for the purpose of developing the industry standard only. Similarly, the business glossary and any data models provided by IBM and other carriers are to be used for the same purpose. It is anticipated that reader of this document isfamiliar with general information modeling concepts.
Figure 1 An overview of the P&C standard development process
2.Approach to developing the P&C Information Models
2.1.Available artifacts (models, glossaries…)
The following artifacts have been made available to the P&C submission team.
- IBM IAA Glossary
- Penn National’s Dimensional model and dictionary
- General Accident’s dictionary
2.2.Available tools
- ER Studio from Embarcadero for conceptual, logical and dimensional modeling
- Unisys Rules Modeler for adding additional metadata/rules to the business glossary
- Adaptive metadata repository to manage business glossary, all the models, traceability and ontology
2.3.Roles and Responsibilities
Role / Responsibilities / Additional Comments/NotesProject Sponsor /
- Vision and scope of the P&C standard
- Promote the development of standard, socialize/recruit P&C SMEs from carriers
- Work with Project coordinator and working group project leaders to ensure timely development of the deliverables
OMG’s representation/
participation /
- Please note OMG’s staff does not participate in developing any standards. OMG provides a neutralProcess to develop standards. Submission to the OMG RFP (timeline, content etc.) is the responsibility of the submission team. Periodically, the submission team may invite an OMG stakeholder to ensure compliance to the OMG Process and answer any related questions.
Project Coordinator /
- Work with Project Sponsor and WG project leaders to ensure timely development of the deliverables
- Ensure compliance to OMG process for development of standards
- Present the progress at quarterly OMG meetings
- Ensure availability of tools needed by submission team and coordinate training as needed
- Assist with import/export of business glossary, information models metadata into the metadata repository
- Assist with consolidation of the P&C submission document
Working group Leaders /
- Work on assigned subject area with SMEs and data modeler(s) to ensure timelines and deliverables are met (including development of project plan)
- As needed assist with resolution of issues, coordinate/communicate progress at monthly status meeting
- Develop assigned sections of the P&C Submission document
- WG leaders for:
- P&C reference model/subject area
- Claim subject area
- Underwriting subject area
- Product management subject area
P&C SMEs /
- Review and validate P&C business glossary as part of the conceptual model validation
- Review and validatelogical and dimensional models (by working with subject area leaders and data modelers)
Data modelers /
- Develop Conceptual, Logical and dimensional models
- Review and update models as needed
Ontologist /
- Develop P&C Ontology that corresponds to the subject areas identified in this approach document (details TBD)
2.4.Timeline
- Proposed start date is June 1st 2008 with final submission to OMG is Q3 2009 (see below).
- Detailed timeline/project plan to be developed by respective working Group Project leaders.
- Proposed date for ‘initial Submission document’ containing (Draft Conceptual model - reference subject area/core model and Claim for example) is September 2008. This timeline (along with other milestones etc.) needs to be confirmed at the May 2nd 2008 Submitters meeting.
2.5.P&C deliverables
This section summarizes the P&C deliverables as specified in the P&C RFP.
2.5.1.P&C Business Glossary
A base version of the IBM IAA Glossary will be imported into the Adaptive metadata repository. This will allow an export/import capability of selected Business Terms from/to the repository. Use of metadata repository will allow different tools and users to access the base glossary as well as ability to classify Terms in different ‘Business Perspectives’ (including subject areas). See Figure 2 for an overview of P&C Business Glossary management. The repository will also allow capturing context specific variations such as ‘Limit’ (whether it is Bodily Injury Limit or rental car allowance limit…).Additional metadata about Business Terms willbe captured and maintained in the repository. This will cover: where used (across various Information models), Source, Format, Author, Status (Draft, Approved, and Deprecated), Synonyms, Version (not at Business Term level). In addition to base business glossary additional context specific metadata including relevant rules (not specific to a carrier) will be captured by importing relevant Business Terms into Unisys Rules Modeler.
Figure 2 Management of P&C Business Glossary
P&C Business Glossary review and validation
The IBM Business glossary contains about 3000 Terms. In order to manage the modeling effort, a set of subject areas mapped to Insurance Business functions will be identified (initial list in section 2.5.2). Relevant business glossary Terms will be used in these subject areas for entity and attribute definitions as applicable. Business experts for each subject area will review and validate the definitions to be used for development of data models. A project leader will be assigned to manage completion of each subject area including resolution of any conflicting and/or missing business Terms.
2.5.2.P&C Information models
Working with P&C SMEs Claim, Underwriting and Product management subject areas have been identified as high priority. This list needs to be validated and agreed upon on May 2nd meeting of the submitters. The following sections list the information models requested in the OMG P&C RFP.
- Conceptual data model
A conceptual data model (CDM) will help SMEs and data modelers identify what entities will be needed to cover the subject areas indicated above. CDM will also contain unique identifiers and relationships among entities as appropriate. For each subject area commonly used entities (such as Party Roles, Location…) will simply refer to the ‘reference subject area’ (see Figure 3). For details on features of CDM please refer to the OMG P&C RFP.
- Logical data model
Conceptual data model for each subject area will be further developed into a fully normalized logical data model with attributes and logical data types and unique identifiers and nature of relationships among entities. Please note that there is no plan to deliver a Physical data model as the RDBMS platform may vary form company to company. For details on features of CDM please refer to the OMG P&C RFP.
- Dimensional model
Dimensional model(s) to support the analytical needs of the carriers will be developed for the high priority subject areas listed above. Please note that the granularity of ‘Fact Tables and need for conformed Dimensions’ will be determined by SMEs to support a standard set of reports that may need to be developed by most carriers. A set of Dimensions (such as Time, Geography…) will use the entities defined in the Reference Subject area but organizes such that they are used in the dimensional model to support analytics.
- Reference subject area
This Subject area will be defined such that it can be used in non-P&C domains of Insurance. A conceptual and logical data model for reference subject area will be developed. Entities included in this area will include (but not limited to) Party, Party Roles (Agent, Claimant, Producer, Adjuster, Insured…), Address, Location, Look-ups/Code lists. Please refer to Fig. 3 for an example.
Figure 3 Example of the ‘reference model/ subject area’ (Conceptual level, not a complete list of entities)
2.5.3.P&C Ontology
A P&C Ontology conforming to OMG’s Ontology Definition Metamodel ( standard will be developed. This Ontology will correspond to the subject areas identified in section 2.3.2. The output format will include XMI and OWL (Ontology Web Language). For details on features of the P&C Ontology please refer to the P&C RFP.
2.5.4.Traceability map
In order to assist data modelers and application developers (and other interested stakeholders) a Traceability map of P&C business concepts to their corresponding models and model elements (including the P&C Ontology and other outputs like DDL, XMI and XML Schemas as appropriate) will be developed. This traceability map will also point to valid semantic variations of standard P&C business Terms.
2.6.P&C business function to process to data mapping
The following diagram depicts how a given P&C business function such as Claim will map to a business process and required data and corresponding information models. Please note that this example is a sample representation of the overall mapping concept and not a complete Claim conceptual and logical model.
Figure 4 Mapping of Claims management business function to Process and Data
Figure 5 Example Claim Subject area (Conceptual level -not a complete list of entities)
Figure 6 Example logical model (snippet only) of Claim Notification
Page 1 of 11 / 10/17/20187:07:33 AM