MindNet is a knowledge representation project that uses our broad-coverage parser to build semantic networks from dictionaries, encyclopedias, and free text. MindNets are produced by a fully automatic process that takes the input text, sentence-breaks it, parses each sentence to build a semantic dependency graph (Logical Form), aggregates these individual graphs into a single large graph, and then assigns probabilistic weights to subgraphs based on their frequency in the corpus as a whole. The project also encompasses a number of mechanisms for searching, sorting, and measuring the similarity of paths in a MindNet. We believe that automatic procedures such as MindNets provide the only credible prospect for acquiring world knowledge on the scale needed to support common-sense reasoning.
MindNet Browsing Now Available! If you are interested in more detailed information about MindNets, a small number of sample MindNets have been made available for online browsing at the mnex project homepage.