BA Insight integrates NLP with AutoClassifier 5.0

Enterprise search developer BA Insight has announced availability of AutoClassifier 5.0, which includes Natural Language Processing to automate extracting key information from documents. The information is then used to generate summaries and categorize and tag all of your enterprise information, including images and videos.

This automation decreases reliance on taxonomies, and when taxonomies are used, AutoClassifier 5.0 employs Machine Learning for taxonomy training to fully automate the tagging process.

When used in conjunction with SmartHub and the company’s 70+ connectors, the combined products provide a complete AI-driven and search engine independent enterprise search solution for organizations of all sizes.

Natural Language Processing includes:

  • Entity Extraction. Automatically generates a list of key ideas to be used for search refinement and to identify similar documents.
  • Document Summary Generation. Automatically summarizes all types of unstructured content so that users can quickly understand the meaning of search results.
  • Sentiment Analysis. Detects the tone in which text is written to determine if it is positive, negative, or neutral.
  • Language Detection. Tags content with the language in which it is written.

For Taxonomy Training, Machine Learning is used to analyse sample documents associated with taxonomy nodes and create training sets, which are used to automatically tag content. Machine Learning-based tagging can be used either in conjunction with or in place of Rules-based tagging so that it can be optimized for each use case.

Image and Video Tagging now includes:

  • Automatic Tagging. Automatic tagging and classification of media content.
  • Transcript Generation. Video transcript generation converts speech within videos into searchable text.
  • Image Extraction. Extraction of images and video from within documents for analysis so that embedded media can be analyzed and tagged.

Integration with Microsoft Cognitive Services, Google Cloud Machine Learning and the Rasa Stack is used to provide intelligence and internet-like capabilities to intranets.

AutoClassifier 5.0 is extensible to allow automatically generated metadata to be stored in many enterprise systems. Currently, plug-in support is provided for SharePoint On-premise, SharePoint Online, ClarityNow, and Elasticsearch.

“By supporting AI capabilities from Microsoft, Google, and the Rasa Stack, we provide our customers the freedom of choice to use the AI platform that best suits their requirements,” said Massood Zarrabian, CEO of BA Insight. “In addition, our AI and rules-based approach gives them the flexibility they need to automate and control the tagging process.”