Semantic clustering

摘要

Semantic clustering techniques are described. In various implementations, a conversational agent is configured to perform semantic clustering of a corpus of user utterances. Semantic clustering may be used to provide a variety of functionality, such as to group a corpus of utterances into semantic clusters in which each cluster pertains to a similar topic. These clusters may then be leveraged to identify topics and assess their relative importance, as for example to prioritize topics whose handling by the conversation agent should be improved. A variety of utterances may be processed using these techniques, such as spoken words, textual descriptions entered via live chat, instant messaging, a website interface, email, SMS, a social network, a blogging or micro-blogging interface, and so on.

主权项

1. A method implemented by one or more computer processors configured to perform semantic clustering, the method comprising:
configuring the one or more computer processors to computationally perform semantic clustering of semantic graphs of user utterances that are from a dialog with one or more executing automated conversational agents by:
computationally producing a respective semantic graph for each of a plurality of user utterances based on linguistic analysis of a corpus that includes the user utterances, the semantic graphs made of one or more nodes and one or more edges, the nodes representing one or more respective concepts within each respective utterance, wherein the nodes represent the concepts abstracted over lexical variations, and the edges representing one or more respective semantic roles within each respective utterance;computationally clustering the semantic graphs of user utterances into a plurality of semantic clusters through use of one or more proximity metrics; andmodifying an operation of the one or more executing automated conversational agents based upon the clustering.