Yahoo! JAPAN R&D

In collaboration with Associate Professor Tadashi Okoshi of Keio Research Institute at SFC, Keio University, the "Mood Deviation Value," which estimates the ups and downs of the user's mood based on user search keywords and other information, was released.

Anatagβ: https://anatag.yahoo.co.jp/

Based on search data, sensor data, and survey questionnaire data obtained with the consent of monitors who cooperated with us in a demonstration experiment in FY 2019, Yahoo! JAPAN Research, in collaboration with the Keio Research Institute at SFC, Keio University, has developed an AI (artificial intelligence) that uses machine learning to estimate moods from search keywords. Based on the search data, sensor data, and survey questionnaires obtained from the monitors, we created an AI (artificial intelligence) "search mood model" in collaboration with Keio University SFC Research Institute. Analysis using the model revealed that there is a relationship between specific search keywords and user mood, that "mood increases on Saturdays and Sundays and decreases on Mondays (after holidays)," that "mood decreases when a state of emergency is declared due to the new corona," and that "mood increases with the first wave, second wave, and so on of the new corona. The mood decreases with the passage of time (presumably because people are becoming accustomed to the new type of corona).

Until now, the Mood Deviation Value has not been available to the public, but in order to let more users know about the fun and future possibilities of AI (Artificial Intelligence) and data-based technology, the Mood Deviation Value are available on the "Anatagβ" page.