Multimedia Contents Retrieval based on 12-Mood Vector
Chang-bae Moon, Jong Yeol Lee, Dong‐Seong Kim, Byeong Man Kim · 2021
The preferences of Web information purchasers are changing. Cost-effectiveness is becoming less regarded than costsatisfaction, which emphasizes the purchaser's psychological satisfaction. In applications of SNS(Social Network Services) based on folksonomy, a method to improve a user's cost-satisfaction in multimedia content retrieval is to use the mood inherent in multimedia items but applications of SNS encounter problems due to synonyms. In our previous study, some problems of synonyms could be solved by using internal tags consisted of arousal and valence (AV) in Thayer's Two-dimensional Model. However, in recall level 0.1, the retrieval performance of the previous study was less than a keyword-based method. In this paper, for improving the retrieval performance of recall level 0.1, a new method using 12 moods vector is proposed, and the proposed method shows good retrieval performance than the previous method and the keyword-based method.