Kyoto University has started joint development with the Dissatisfaction Purchase Center Co., Ltd., which manages consumer dissatisfaction survey data, with the aim of "applying the latest natural language processing technology to business."
Sadao Kurohashi, a professor at the Graduate School of Informatics, Kyoto University, who is involved in this research, has been studying natural language processing and knowledge information processing for many years. We have received numerous awards.Under such circumstances, in order to realize the application of the latest technology to business, the Kurohashi / Kawahara Laboratory of Kyoto University is the CREST project of the Strategic Creative Research Promotion Project of the National Research and Development Corporation (JST) "Knowledge". Improvement of marketing support service by introducing the robust text analysis technology and predicate term structure analysis technology, which are being researched and developed in "Establishment of structural language processing and construction of knowledge infrastructure" in "Dissatisfaction Survey Report". You can improve efficiency.
The "Dissatisfaction Survey Report" is a "dissatisfaction investigation report" sold by the Dissatisfaction Purchase Center, which "purchases dissatisfactions occurring in each industry or company from consumers and analyzes the data according to the product name / company name and consumer attributes. , Marketing support for companies. "Until now, it has been created by analyzing the data collected by LDA technology (potential Dirichlet allocation method), but although it is possible to classify texts with LDA, it is possible to judge the classification criteria of texts determined mechanically. Needed to be interpreted by an analyst, and it took a lot of effort.
In addition, various researches have taken the lead in the fields of big data and data analysis, and while it has attracted public attention, it has been said that the diversion of natural language processing technology to business value has not progressed so much.
Through this joint research, the improvement and efficiency of marketing support services have been realized.In addition, by applying the latest natural language processing technology, it is possible not only to make summarization based on the structure of sentences and improve analysis accuracy, but also to automatically output results in a form that is easy for humans to understand. It is now possible.