A Study of the Correlation Between MBTI and Social Media Food Recommendations

Shutong Yang

DOI: https://doi.org/10.54254/2753-7064/19/20231199

2023-12-07

Abstract:With the booming of social media, diverse and accurate user figures can significantly improve personalized experiences in industrial applications like advertisements and recommender systems. Meanwhile, deep learning is widely used in user tag construction, which means mining labels with practical significance through historical interaction behavior. For instance, clicking on advertisements and conversions can express the tags of users interests and intentions. The construction of accurate user figures is difficult, the potential of existing modal data has been plumbed, and the challenge is the need for further features. Then, optimizing recommender systems accuracy through diverse methods becomes the key to the sustainable development of each social media platform. Recently, MyersBriggs Type Indicator (MBTI) test results become a highly discussed topic in the mass media, and enough samples have been accumulated to confirm its scientific validity. Indeed, MBTI results strongly correlate with personal habits and preferences, which the recommender system of social media has not fully applied. Therefore, this paper has researched optimizing the recommender system of social media platform content based on MBTI results, which mainly focus on the food topic of the RED app. The following has adopted the questionnaire as the primary research method, supplemented by the literature review and observation method. The research concluded that different MBTI have different dining habits and preferences. As preliminary research of relevance, the value of this discovery can help develop accurate user profiles and optimize the quality of content recommendations.