Extracting useful information from Facebook posts: A Cognitive-Linguistic Analysis

Abstract

�Social media has emerged as a significant communication platform in recent times where millions of people across the world share thoughts and ideas and interact with each other via social networking sites. Researchers examining communication in social networking sites have stated that communication in these forums has a strong phatic (small-talk) element and there is very little substantive information content in posts. This paper re-examines the view and suggests that posts in social networking sites is not always phatic. We argue that information content of a certain �post� needs to be examined from the perspective of both the sender and the receiver of information. For the purpose of our study, we conducted an empirical study, where users rated posts on their Facebook wall on the basis of their �information content� and �phatic element�. Thereafter we undertake a content analysis of posts using tools from pragmatics and demonstrate how users extract implicit information from the content of these posts. We conclude that a detailed and in-depth linguistic analysis based on principal of pragmatics needs to be undertaken to design algorithms for automatic identification of meaningful posts from �apparently phatic� social media posts that would help in making useful organizational decisions.