Abstract— As the internet environment evolves and new media emerge, consumers start to share their opinions and reviews of products on the web. There is also growing demands for analyzing such online reviews and identifying consumers’ true minds to meet these emerging trends while a large number of studies have been made on online reviews in a wide range of academic fields including marketing, MIS and computer science. However there has been little research conducted on video game industries dealing with typical experiential products. Thus, this study was intended to analyze community data in games domain available on STEAM, a world-wide game platform, using a data mining approach. Several machine learning techniques such as Classification and Regression Tree (CART), Artificial Neural Network (ANN) were applied to community data collected from STEAM games to analyze factors that have impact on helpfulness of game reviews. We also conduct sentiment analysis of review comments to mashup sentiment results to original data set. We will provide analysis results and interpretation of the results with further research directions.
Index Terms— Community of game users, data mining, online reviews, STEAM, usefulness of reviews.
Ha-Na Kang and Hye-Ryeon Yong are with Graduate School of Interaction Design, Hallym University, Chun-Cheon, Republic of Korea (e-mail: khnnn0607@naver.com, yong-@naver.com).
Hyun-Seok Hwang is with the Business Administration Department, Hallym University, Chun-Cheon, Republic of Korea (e-mail: hshwang@hallym.ac.kr).
[PDF]
Cite: Ha-Na Kang, Hye-Ryeon Yong, and Hyun-Seok Hwang, " A Study of Analyzing on Online Game Reviews using a Data Mining Approach: STEAM Community Data," International Journal of Innovation, Management and Technology vol. 8, no. 2, pp. 90-94, 2017.