data-language="eng" data-ev-field="abstract">Maintaining weight loss is known to enhance the long-term health condition for obese people and require an effective strategy to implement. This paper is aimed at exploring ways for obese people to maintain weight after weight loss by data mining technology, highlighting the association between social problem solving and weight loss outcomes. The research is achieved by using the Social Problem Solving Scale and collecting questionnaires and weight data from 101 team members who have left the online fat loss class for three months, where 27 of them are interviewed online by voice or video. The SPSS 26.0 is applied to test the predictive effect of social problem solving on weight loss. The results reveal that the weight loss effects after three months is significantly associated with negative problem tendency (γ=-0.23, P<0.05). According to regression analysis, negative problem solving tendency can accurately predict weight for the people losing weight, which is successfully validated by statistical significance with effect value. Our research concludes that an improvement on individual's social problem solving skills can lead to a remarkable transformation from negative and other irrational problem solving to positive and rational problem solving. This can increasingly facilitate health benefits, such as an effective control on weight and a significant decrease on the potential for various diseases, subsequently offering the sense of well-being and life satisfaction for individuals.