Semantic MediaWiki Approach to Metadata
Are semantic technologies a utopic vision or natural paradigm shift? How can we use semantic technologies to collect metadata and enable the semantic search capabilities of the future?
The volumes of data, rates of change, and levels of complexity continue to increase in the financial industry making it even more challenging to find the data we need. Add to this the lack of a common language and you end up with “the last mile problem”. Projects produce valuable data assets but fail to effectively communicate the new data inventory to the enterprise (no enterprise architecture). The resources that worked on the projects move on to other things and the collective learning is lost, even though cryptic documentation is stored away in repositories never to be seen again. This presentation examines using Semantic MediaWiki to begin building a knowledgebase with an underlying structured ontology:
- Leveraging semantic technologies to improve data find-ability
- Making metadata everyone’s job responsibility
- Capabilities and processes to access the collective knowledge of the organization
- Collective learning produces shared understanding
- It’s not about controlling change it’s about working at the rate of change
- Think globally and act locally
Scott E. Thompson is the manager of data architecture at Ontario Teachers’ Pension Plan, the largest single-profession pension plan in Canada, and the programme director of IRMAC, the oldest DAMA chapter in North America. Scott has over 12 years experience in software development, project management, and data management areas, including architecture design, implementation, and data quality management of business intelligence solutions. Scott is currently focused on enterprise solutions in the financial services industry and expanding the role of data as a strategic enabler of competitive advantage and corporate objectives.
How to Get Business Value from Metadata
Katrina Easton and John Tench will describe an innovative Metadata Mapper that was implemented at Intact Financial Corporation. The solution provides an intuitive picture of a highly complex network of data flows and data dictionaries that has proven appealing to business and technology users. Unique features include:
- End-to-end data mapping
- Linkage of data definitions to data profiling results
- Hot links to code tables
- Enterprise Data Model cross referencing
- Metadata workflow support
- Customisable loading of ETL mapping spreadsheets and other BI artefacts
The IRMAC membership database has recently been migrated from Microsoft Access to a web implementation based on MySQL. Some of the learnings and tools used with Metadata Mapper have been leveraged to provide an integrated data dictionary. John Tench will explain how this was done and will demo EMU, an Excel-based utility for the exchange of information with popular modelling tools and formats.