Mining of marital distress from microblogging social networks: A case study on Sina Weibo

Marital distress, occurring when a married person encounters a profound sense of disappointment or doubts about staying married, has been regarded as the causes of many risks, including child behavior, economic stability and problems with mental and physical health. Therefore, it is important to discover the people with marital distress and take proactive measures accordingly. Traditional approaches of discovering marital distress are subjective by using self-reports, interviews and questionnaires, which are limited to data scale and error-prone in the early stage. In the era of big data, it is easy to apply efficient machine learning based approaches on massive social data. Therefore, we propose a novel model, named Discovering Marital Distress (DMD), to discover the crowds with marital distress, where there are four components, i.e., Personal Profile, Daily Posting Habits, Interactive Behaviors and Emotion Measure. The experimental results demonstrate that the proposed DMD model is effective on the issue of finding persons with marital distress, with precision of 96.5% and recall of 96.4%. And through contrast experiments, it is found that the persons with marital distress update their micro-blog more frequently and mostly at dead of night, and the content of these posts are long and negative almost without pictures.