Assessing Trustworthiness in Social Media

A Social Computing Approach

Project Description

Social media is gaining popularity in recent years and increasingly becoming an integral part of our life. Given the extensiveness, instantaneity, and diffusion speed of social media, e.g., a tweet or a clip of video, could galvanize a digital revolution or wreak havoc with one's otherwise routine and uneventful working life. With the presence of adversaries, the convenient use of and low barrier of social media brings about new challenges. How well we address these challenges can directly influence our ability to manage information and misinformation, and the future role of social media as a reliable communication mechanism. One such pressing challenge is to assess information trustworthiness in social media . We propose to investigate research issues related to social media trustworthiness and its assessment by leveraging social research methods, developing new computational social methods, and creating novel approaches to social media data collection and sharing.

Research Problem

In social sciences, trust is about a relationship between two entities, the trustor and the trustee. Trust can be defined as the perception of the trustor about the degree to which the trustee would satisfy an expectation. Trustworthiness can be defined from the perspective of both entities; in this work, it is the perspective of the trustor that defines a property that can be judged, i.e., the amount of trust associated with the trustee. In all cases trust is a heuristic decision rule, allowing the human to deal with complexities that would require unrealistic effort in rational reasoning. One of the key current challenge is to rethink how the rapid progress of technology has impacted trust as information technology has significantly changed how people interact, express themselves, and behave. The assessment of information trustworthiness in social media requires answers to the three essential questions about the information: (1) source (or author), (2) author position, and (3) content. The search for the answers is greatly complicated by the nature of social media: enormous sizes in terms of users and links, irregular uses of languages, incomplete sentences or messages, and inordinate amounts of data and meta data. In addition, both linked data and attribute data are present in social media. The former represents the connectedness among entities and the latter the properties of entities. In search of the three answers, we face research challenges:

  1. Information Provenance - Identifying the true source (or author) of information,
  2. Friendship Differentiation - Determining if the author is a friend, acquaintance, or foe, and
  3. Content Analysis - Analyzing the content to ascertain its intention, quality, and etc.
In this project, we focus on developing computational social theories and methods for the first two challenges. The third challenge is partially addressed in our recent work. Additional work on trust maintenance can be found in literature.

Subject Terms

Social Networks, Social Media, Social Media Mining, Trust, Information Provenance.



Related News

Interdisciplinary Collaboration

Project Members (current and former)


This material is based upon work supported by, or in part by, the U.S. Army Research Office (ARO) under contract/grant number 025071. Any opinions, findings, and conclusions or recommendations expressed here are those of the author(s) and do not necessarily reflect the views of ARO.

Created by Huan Liu who can be reached at
Webmasters: Pritam Gundecha, Email:;
                    Jiliang Tang, Email:

Last Updated: April 14, 2015