Numerous inquiries about drugs are received at the call centers of pharmaceutical companies. Since the nature and content of inquiries are diverse, accurate responses will be difficult if there is no experience in preparing for the launch of a new product. This study focused on the versatility of information available for products in the early stage of the product launch. Inquiry contents in the early stage of the product launch as input data were extracted as concepts by text analysis. Then, the extracted concepts were structured and visualized. The extracted concepts were applied to the data of the same product with different periods and the data of a different product with different disease areas, to verify the versatility. As a result, it was confirmed that the extracted concepts from one product can be comprehensively applied to the period after the launch, and to the product having different disease area.

A general QA system realizes automatic answering by analyzing questions given by the user based on a model that retrieves appropriate answer candidates for each question from a set of documents (an answering documents retrieval model). When constructing an answering documents retrieval model for question/answer documents for a community QA (Question Answering) sites (cQA sites), a method based on a similarity measure on question documents can be a basic method. However, it is difficult to present appropriate answering documents while accurately grasping the diversity of answer documents that exist for each question as seen on a cQA site. In this study, we propose a method for constructing an answering documents retrieval model that can consider the diversity of answer documents by using a topic model. In order to verify the effectiveness of a proposed model, a verification experiment using question/answer documents actually posted on a cQA site is performed.

It is an important and challenging issue to determine an appropriate exhibition price for each item on a second-hand fashion EC (Electronic commerce) site. If an item is not sold for a certain period, it should reduce the price of the item. Because second-hand fashion items have various features such as brand and category, it is not easy to accurately estimate the selling prices of each item that would be accepted by a customer. In this study, we propose a model that analyzes the relationship between the first exhibition price and the actual selling price of an item by analyzing its past sales history data and the results of the price change test in an integrated manner. To obtain a high prediction accuracy, items were clustered based on their characteristics and prediction accuracy, and the result of the price change test was used to verify the response of each customer. Using actual data analysis, we identified the items with a high prediction accuracy and predicted the selling price when the first exhibition price was changed for those items. Furthermore, we analyzed the appropriate setting of the selling price for future.