Discover the breakthrough insights that lead to new business opportunities and reveal significant lessons learned, not just delivering search results or finding things that you already know.
WHAT IS IT:
Business Insights Workbench (BIW) for Information On Demand is an information analytics technology that derives critical business insights by mining structured and unstructured information. It also combines traditional Business Intelligence techniques with text mining techniques in a single platform to provide a full range of analytics capabilities. The technology allows business users to move away from endless googling and focus on discovering valuable business insights.
HOW IT WORKS:
To discover business insights from an ocean of information, BIW incorporates interactive information analytics methodologies including clustering, taxonomy generation, co-occurrence analytics, classification, annotation, trending, and visualization to derive values out of vast amount of information. In addition, BIW contains a set of data warehousing capabilities to allow users to build scalable data warehouses that host structured and unstructured information. Structured information provides context to the unstructured information, such as authors, time of the document creation, etc. It is crucial to link these two kinds of information to derive insights.
WHY IT’S UNIQUE:
BIW extracts insights from free form text which represents 65% of all data kept by a company, and integrates it with structured data about customers, performance, and products. The results of these efforts have been the development of a number of solutions addressing the professional’s needs in the areas of:
- “Detecting Early Opportunities and Threats from Customer content”
- “Analyzing Intellectual Property Portfolio for Infringement and Opportunity Analysis”
- “Early Warning Identification of Product Issues”
- “Tracking and Forecasting Technology Maturity”
- “Discovering Insights into the evolution of an Emerging Market”
BIW can leverage IBM OminFind technology for full text search and content extraction. The primary methodology employed is the creation of natural classifications (taxonomies) of data objects. Experience has shown that the only way to insure the “naturalness” of taxonomies is through expert human intervention at every stage of the taxonomy generation and modeling process. Once the taxonomy is complete, it is treated as another dimension in the database and analyzed against other taxonomies or in conjunction with structured information to find interesting trends or co-occurrences.
Business Insights Workbench real-time data manipulation gives the professional and extensive range of capabilities, including:
- Drill down to any level of detail
- Analyze only parts of a document
- Create customized affinity analysis with multiple views
- Compare and merge analysis from different sources or previous analysis
- Pattern detection using multi-dimensional affinity analysis
- Scatter charts automatically identifies non-obvious relationships
- Automatic extraction for up to 100 different types of entities
- Automatically detect "knowledge gaps" in a set of text suggesting areas for further analysis