According to a survey, approximately 70% of the chance of marrige potentialy exists in local networks, such as a workspace and school, and also the possibility of marrigement with a partner found through the Internet is likely low. Existing matching services are designed that people can find out their new partner without being revealed to use the services by their friends. Thus, we tried to develop a service in which people can match with compatible close friends and collegues without any awkward, and we assume that the service help to solve the low birth rate problem. One of the main features of the service is to let communication with your partner activate with Natural Language Processing and Machine Learning; namely, users of the service are suggested some topics which interest your partner by the system so that they can communicate more smoothly. (This work was a team project and I was in charge of the recommandation system with machine learning and natural language processing.)

Contest: JPHACKS2017 - Type: Web Application