Kato, M. P. Laboratory
In Kato Laboratory (Knowledge Acquisition Systems Laboratory), we have been conducting joint research with Yahoo! JAPAN Research of Yahoo Japan Corporation, and have been studying search algorithms, their evaluation techniques, and analysis of search user behavior. One of our recent achievements is the demonstration of a search algorithm evaluation technique in Yahoo! Chiebukuro. We used a technique called “interleaving”, which is said to be far more efficient than conventional evaluation techniques, to evaluate search algorithms in Yahoo! Chiebukuro for six months. If we can efficiently find good search algorithms by using these techniques, it will be easier to find necessary information in Web services such as Yahoo! Chiebukuro. In 2021, Yuki Yanagida, who belongs to the master’s program, is analyzing both search behavior in Yahoo! Search and purchase history in Yahoo! Shopping to answer the question of what kinds of searches make people more satisfied with their purchases. Once it is revealed, we can apply it to a variety of applications, such as predicting purchase satisfaction and supporting pre-purchase searches. Our laboratory is focusing on a general retrieval model for structured data, which can be used for product search and landmark search, and recruiting students who are interested in search technologies.