Location emotion recognition for travel recommendation based on social network

With the rapid extension of social network, a huge amount of text and multimedia data concerning venues have constantly been generated. More and more people now try to obtain others’ reviews on one venue from the different social network places. Thus, it is essential to analyze both text and multimedia content in an integral manner to get a better venue semantic evaluation, which can effectively be applied to several applications such as travel recommendation, venue summarization, emergency monitoring. In this paper, we propose a novel multimedia location emotion recognition model to handle this problem. First, we utilize traditional and classic emotion multimedia datasets to train several recognition models according to different modals. Then, we propose a novel method to fuse the recognition results provided by these different pre-train recognition models and output the emotion label of the given venue. Finally, we recommend the related venues to users with respect to the emotion label. In order to demonstrate the performance of our approach, we collect the related location multimedia data. The related experiments also demonstrate the superiority of our approach.