The most popular and comprehensive Open Source ECM platform
Let us know more
Business Intelligence: Assisted by Enterprise Search for Unified Information Access
Business Intelligence (BI) is increasingly being used by businesses for synthesizing data culled from many company databases to create reports and dashboards that can identify patterns and trends. BI has had its share of success, but it sometimes isn’t able to tell the whole story. That’s because BI is limited to analyzing structured data stored in databases. It doesn’t consider unstructured data.
But the amount of unstructured data is growing much more quickly than the amount of structured data. For example, there are significant sources of unstructured data being stored in organizations, particularly in documents and emails that are basically untapped by BI. That’s not to mention information being created on blogs, Facebook pages, wikis and microblogging platforms like Twitter. A typical ratio of unstructured to structured data found in an organization is about 80:20. Unstructured data isn’t easy to analyze, and because of that, unstructured data hasn’t typically been included in the analysis of BI tools.
Unlike BI, Enterprise Search focuses on trying to uncover information held in unstructured data, but there are limitations as to what search can do. While search can help identify unstructured content, it isn’t very good in analyzing what the content means. BI is based on analytics that have been built on top of the structured search SQL queries, but not much work has been done to create a similar sort of analytic framework to run on top of unstructured search results.
While analysis of unstructured data still remains a stumbling block, recently vendors are trying to at least unite the worlds of BI and Enterprise search to provide a more comprehensive approach for finding business information and understanding important trends and patterns in that data. The result is something that is being called Unified Information Access (UIA). UIA is a term that is being popularized by Forrester, and some vendors, like Attivio, are beginning to target products in this space.
Most Enterprise Search solutions try to mimic the simple one-field search method popularized by Google and are more recently also including faceted search as an option. On the other hand, BI applications tend to have much more complex interfaces for specifying analysis criteria, often requiring many pages of options to be completed. UIA tries to combine the two technologies. UIA queries are built as a combination of both search terms and analysis criteria. The results are then returned in a format that resembles a portal or mashup. In this way, the results from the two technologies are combined and reported on within the same result screen. One of the interesting aspects of UIA is that analytic results are supplemented with links into structured data elements.
Using UIA, analytic results are improved because unstructured data can be brought in as backup information to support the conclusions of the analytics. The approach allows a user to be able to interact with all data in the corporation, not just the unstructured or just the structured data components.
UIA is an interesting idea and I’d expect that we will see more BI products try to integrate unstructured data into their analytic results.
