Questioner or question: Predicting the response rate in social question and answering on Sina Weibo

•Address the problem of low response rate in social Q&A that happens on a Chinese microblogging site.•Propose 17 factors from both the questioner and the question's perspectives and investigate their effectiveness in determine whether or not a question will be answered.•Build a model to differentiate questions with high probabilities of being responded from those with low probabilities of being answered, which achieved a prediction accuracy of 74%.•Our results show that whether or not a question will be answered depend more on the questioner than the question.•Propose the potential implication of improving visibility of users with low social capital to increase their question's response probability. With the noted popularity of social networking sites, people increasingly rely on these social networks to address their information needs. Although social question and answering is potentially an important venue seeking information online, it, unfortunately, suffers from a problem of low response rate, with the majority of questions receiving no response. To understand why the response rate of social question and answering is low and hopefully to increase it in the future, this research analyzes extrinsic factors that may influence the response probability of questions posted on Sina Weibo. We propose 17 influential factors from 2 different perspectives: the content of the question, and the characteristics of the questioner. We also train a prediction model to forecast a question's likelihood of being responded based on the proposed features We test our predictive model on more than 60,000 real-world questions posted on Weibo, which generate more than 600,000 responses. Findings show that a Weibo's question answerability is primarily contingent on the questioner versus the question. Our findings indicate that using appreciation emojis can increase a question's response probability, whereas the use of hashtags negatively influences the chances of receiving answers. Our contribution is in providing insights for the design and development of future social question and answering tools, as well as for enhancing social network users’ collaboration in supporting social information seeking activities.