Autonomy, acquired by HP in 2011, established a range of analytics tools, that converge information from many different information sources and let the user explore the synthesised view (“exploratory analytics“). Initially, Autonomy established their technology in statistical pattern recognition techniques (Bayesian inference) and later on expanded into enterprise search and knowledge management applications. Finally, the technology portfolio was structured into IDOL (Intelligent Data Operating Layer).
In comparison to IBM Watson and CMU NELL, HP Autonomy IDOL seems to have the broadest range of paying customers. Since HP Autonomy IDOL seems to be lacking a knowledge graph as a technical component, it seems to be less “thinking” (i.e. deriving new knowledge automatically from deep interpretation of data) but actually more “analysing” data to derive descriptive, bird’s-eye-persepctives.
Recently, it seems that HP is planning to market Autonomy components as single modules within a PaaS paradigm. This would move it further away from Watson towards startups like Alchemy.
The Cognitive Workbench cannot cover all areas that HP Autonomy IDOL covers, mainly because of IDOLs picture and video processing capabilities (ExB is developing such technologies outside the Cognitive Workbench paradigm). For the rest, the ingredients-list seems to be:
- Data importers (Facebook, Twitter, .doc, etc)
- Named Entity Recognition
- Time Extraction
- Keyword Extraction