Quantitative analysis of emotionally expressive language for personalized service

Researches on personalization services with context awareness under a mobile environment are being conducted actively. Majority of them focus on user profile-based learning which is in need of personalization or context-awareness-based reasoning technique and does not cover the emotional aspect of users. In order to sense emotion information, which may not easily be sensed from users, this research analyzed the tendency of positive/negative aspects of emotional expressive language, which was sorted from a psychological emotion model using AHP. Interrelationship analysis and matching between certain emotional expressive language and cognized context were performed as well. A survey was carried out among 100 persons in their early 20s and the result of the survey was analyzed via AHP. 91.6% of the test group showed that there was strong interrelation between certain context information and positive emotion expressive language. Therefore, an emotional personalization service that uses universal emotion expressive language may be provided for accurate context awareness to be performed.