System and method for automatically mining corpus of communications and identifying actions

摘要

Exemplary embodiments of the present disclosure are directed towards a system and method for automatically mining a corpus of communications and identifying actions. The method comprises collecting corpus of communications from the communication modalities, determining the importance of the senders of those communications, filtering the corpus of communications remove promotional or marketing content, header and footer contents including signatures. The method further includes segmenting the filtered content to define phrases, creating a canonical form for each of those phrases, and extracting a feature vector for the canonical form. The feature vectors are processed by a classification algorithm to determine whether the corresponding phrase is one which requires the user’s attention or not.