Professor Lin’s Analysis on Facebook Big Data to Predict 2016 Presidential Election

Understanding the ideology on social media has been a critical issue focused by politic, communication and economic scholars. It is also one of the core research directions of Taiwan Institute for Governance and Communication Research (TIGCR). 

TIGCR was pleased to invite Ming-Jen Lin, Distinguished Professor and Chair of Economics, National Taiwan University (NTU) to visit our center, and deliver interesting lecture on “Using Facebook Data to Predict the 2016 U.S. Presidential Election” on November 21. He is also one of the Asian representatives of Social Science One (SSO) that he shared with our members about the recent development of SSO in the talk. 

His lecture was mainly derived from his research, in which he analyzed Facebook Big Data from 2015 to 2016 with more than 1.9 billion likes and 2000 presidential election-related fan pages on Facebook. He mentioned that most of past studies of ideology on social media focused on polarization, however, he suggested that suitable measurement can identify the continuity of ideology on social media. In his research, Lin not only concentrated on the politician fan pages, but also included fan pages that related to users daily lives and media they accessed. By principal component analysis (PCA), he was able to guess the certain types of ideology existing on social media, and he confirmed the phenomenon of echo chambers on Facebook based on the research outcome.

Timing is another important dimension applied in Lin’s research. Lin indicated that media ideology dynamic did not fluctuate wildly. Nevertheless, interesting findings situated at ideology on politician fan page. Lin said, some politician stand firm on the political issues all the time, while some change communication strategies in the election periods to make users on fan page not to be polarized.

Lin's lecture

(Photo credit: TIGCR) 

Apart from the in-depth analysis on the Facebook Big Data, Lin also shared with members of TIGCR about the recent development of SSO. “In the long term, I hope maybe SSO could provide a link or let us extract Facebook data under supervision to prevent data leaks,” said by Lin. “As representative, I am not that optimistic. Facebook is under great pressure now. The possible condition to utilize its databases is to promise assembling distinguished scholars with more rigorous method to study. Just wait and see.” In the insightful discussion, Lin awarded high recognition to the current researches conducted by TIGCR. We are also looking forwards to having more corporation chances with Lin to bring academic vitality to TIGCR. 

Group photo after the lecture

(Photo credit: TIGCR)