atomicintelligence^™

Text Processing Technology

Here are some example configurations for various tasks:

• Competitive intelligence
  atomicSearch, atomicClassification, atomicAnalytics
• Extraction of data from diverse incoming content
  atomicParsing, atomicClassification, atomicSearch
• Extraction of data from single web page
  atomicParsing
  
• High-accuracy Intranet search
  Google Search Appliance, atomicClassification

• Rule-based parsing
  Handles both linguistic and tabular content.
• Machine learning
  Automatic learning and optimization of rules.
• Named-entity detection
  Resolves company names by use of context

• Conditional classifiers
  Classifiers are invoked depending on conditions related to search terms, content type, source identity.
• Choice of underlying methodology
  Naive Bayes, Support Vector Machines

• State-of-the-art indexing and search
  Full text, fielded indexing. Scalable. Supports personal and shared indexes.
• Integration with Google
  Integration with Google Search Appliance, Google Mini, and Google Desktop for Enterprise.

• Intra-document analytics
  Significant phrases, buzz terms.
• Inter-document analytics
  Volume, share of volume, ranked relatedness, trends.
• Visualization
  Charts provide a new way to navigate and drill-down into sections of data.