What Makes a Candidate Win? Examining the Effect of Different Topics of Election Promise on Councilor Election Result in Taiwan

Research output: Conference Papers (RGC: 31A, 31B, 32, 33)PosterNot applicablepeer-review

View graph of relations

Author(s)

  • Xin Jin
  • Ziqi Yi
  • Donglin Xu

Related Research Unit(s)

Detail(s)

Original languageEnglish
Publication statusPublished - Aug 2019

Conference

Title8th National Conference on Social Media Processing (SMP 2019)
PlaceChina
CityShenzhen
Period16 - 18 August 2019

Abstract

The result of logistic regression based on the information of 2018 Taiwan city and county level councilor election candidates (N = 1753) indicates that, under the context of Taiwan democratic politics, the election result is significantly affected by election promises of candidates, meaning policy preference indeed exists during the process of Taiwan councilor election. Specifically, #education, #public facilities, #smart city, and other 19 topics mentioned in election promises can positively predict the successful election result. Candidates information was collected on a voter guidance website, and topics of election promises were extracted with the help of word2vec and K-means clustering, two unsupervised machine learning methods.

Research Area(s)

  • Computational Communication Research, Political Communication

Bibliographic Note

Information for this record is supplemented by the author(s) concerned.

Citation Format(s)

What Makes a Candidate Win? Examining the Effect of Different Topics of Election Promise on Councilor Election Result in Taiwan. / Jin, Xin; Yi, Ziqi; Xu, Donglin.

2019. Poster session presented at 8th National Conference on Social Media Processing (SMP 2019), Shenzhen, China.

Research output: Conference Papers (RGC: 31A, 31B, 32, 33)PosterNot applicablepeer-review