TEXT MATCHING DEVICE AND METHOD, AND TEXT CLASSIFICATION DEVICE AND METHOD

摘要

[Object] To provide a system for automatically and reliably collecting information belonging to a given category, and matching the information appropriately in a timely manner.;[Solution] A text classifying device 30 analyzes dependency of collected texts by a morpheme analyzing unit 52 and a dependency analyzing unit 54. A problem report collecting unit 64 specifies a core consisting of noun+predicate in a text based on dependency relation of the text, and using a combination of noun classification (trouble/non-trouble) and predicate classification (excitatory/inhibitory), classifies the text to a problem report or the rest, by a method referred to as core-based matrix. Support information collecting device 66 and request message collecting device 68 collect support information and request messages in the similar manner. A matching device 76 matches problem reports and support information collected by problem report collecting unit 64 and support information collecting device 66 by a method referred to as co-occurrence core matrix, and thus associates problem information (support information) with appropriate support information (problem information).

主权项

1. A text matching device, matching, in a set of texts classified to a first or second category, a text in said first category with a text in said second category, wherein
a text included in said set is classified to said first or second category by a text classifying device using machine learning, using as features, one or a plurality of morphemes forming the text, dependency information of the one or a plurality of morphemes, and a combination of a noun classification and a predicate classification in a core of a sentence consisting of a combination of a noun included in said text and a predicate on which the noun depends; said text matching device comprising: storage means for storing texts of said first category and said second category distinguished from each other; text pair generating means for generating a text pair consisting of a text of said first category and a text of said second category; features-for-matching generating means for generating, from said pair, features-for-matching, including said features used when the text in said pair generated by said text pair generating means is classified by said text classifying device; and matching means for determining, using the features-for-matching generated by said features-for-matching generating means, whether two texts forming said pair match or not; wherein said matching means includes a machine learning model pre-trained using training data for matching in advance, to determine whether a pair of texts matches based on said features-for-matching.