Negative emotions boost user activity at BBC forum

An empirical study of user activity in online BBC discussion forums, measured by the number of posts written by individual debaters and the average sentiment of these posts, shows that most posts contain negative emotions and the most active users in individual threads express predominantly negative sentiments.

Abstract

We present an empirical study of user activity in online BBC discussion\nforums, measured by the number of posts written by individual debaters and the\naverage sentiment of these posts. Nearly 2.5 million posts from over 18\nthousand users were investigated. Scale free distributions were observed for\nactivity in individual discussion threads as well as for overall activity. The\nnumber of unique users in a thread normalized by the thread length decays with\nthread length, suggesting that thread life is sustained by mutual discussions\nrather than by independent comments. Automatic sentiment analysis shows that\nmost posts contain negative emotions and the most active users in individual\nthreads express predominantly negative sentiments. It follows that the average\nemotion of longer threads is more negative and that threads can be sustained by\nnegative comments. An agent based computer simulation model has been used to\nreproduce several essential characteristics of the analyzed system. The model\nstresses the role of discussions between users, especially emotionally laden\nquarrels between supporters of opposite opinions, and represents many observed\nstatistics of the forum.\n