IBM's Watson is a computing system that became famous in 2011 for beating human champions at the televised game Jeopardy. This game, as most of you know, involves answering questions on a variety of subjects. Contestants can win up to one million dollars. Watson fits into this picture as a question and answer system. Where Web search engines take key words and deliver mountains of links to possibly relevant documents a question and answer technology such as Watson aims to parse the question, reason about it, and reply with a precise answer. Although Watson is a proprietary hardware and software platform for deploying question and answer applications if you are adventurous and handy with Java code you can roll your own using open source code. The OAQA project, Open Advancement of Question Answering Systems at https://oaqa.github.io/ is an endeavor that has defined a question and answer architecture to be used for research in the field. They have also built Java libraries useful for constructing question and answer systems. This ongoing effort is a collaboration between the Language Technologies Institute of the School of Computer Science at Carnegie Mellon University and IBM's Deep QA Group http://researcher.watson.ibm.com/researcher/view_group.php?id=2099 . A Carnegie Mellon professor, Eric Nyberg http://www.cs.cmu.edu/~ehn/ , is a key figure in the advancement of this research that is the underpinning of Watson. Nyberg, along with IBM researcher David Ferrucci http://www-03.ibm.com/innovation/us/watson/research-team/dr-david-ferrucci.html , conceived of a question and answer architecture that could be and has been extended far beyond winning at Jeopardy. The project provides their OAQA Tutorial on GitHub https://github.com/oaqa/oaqa-tutorial/wiki/Tutorial along with Java code libraries CSE-Framework https://github.com/oaqa/cse-framework and BaseQA https://github.com/oaqa/baseqa From the tutorial: “Unstructured Information Management applications are software systems that analyze large volumes of unstructured information in order to discover knowledge that is relevant to an end user. An example UIM application might ingest plain text and identify entities, such as persons, places, organizations; or relations, such as works-for or located-at.” I am not suggesting that you try to compete with IBM with their platform of 2880 execution cores running simultaneously. But if you are interested in seeing how Watson gets things done with natural language question and answer it is entirely possible to build a tiny Watson that operates in a limited universe of data.