Open Data Profiling
Generate a profile of people or organisations based on web data sources.
Organisations have a need to understand in more detail entities they deal with as their internal data is not evolving at web speed. Gathering important information from open data on the web is becoming crucial to banks, insurances and other companies to insure they provide the right services in context of that entity. However, gathering facts from text written on the web is a challenge as it is unreliable, incomplete and duplicated.
What we did
By leveraging the Text analytics component of the IBM Big Insights platform we were able to extract potential facts in context. Then, by leveraging other components and creating our own disambiguations engine, we were able to combine these candidates into profiles for further consumption by the business users using Data explorer. We created the solution so that we can apply many types of algorithms and text analytics rules in parrallel to minimise ambiguity and provide a scalable platform at the same time.
Open Data Analytics (ODA) is a framework that can be adapted to many different profile styles and leveraged in many businesses while being very scalable and extensible. It can easily extend the enterprise Master Data Management (MDM) systems.