社群媒體情緒與房市交易資訊關係之檢視

An Examination of the Relationship between Social Media Sentiment and Housing Market Information

朱芳妮(Fang-Ni Chu) ； 高嘉璘(Chia-Lin Kao) ； 洪志興(Chih-Hsing Hung) ； 陳明吉(Ming-Chi Chen)

《應用經濟論叢》 113期 (2023/06) Pp. 85-138

https://doi.org/10.53106/054696002023060113003

摘要

透過社群網路搜尋人們的看法評價作為參考已成為現代人做決策的一種方式，過去文獻已發現傳統媒體是影響市場的重要因素，本文進一步了解社群媒體與房市變化是否有關聯。本文使用OpView社群口碑資料庫2016年至2020年有關房市的臉書（Facebook）與論壇（PTT）貼文，建構社群媒體情緒聲量指標，與房價、交易量、流通天數、議價空間等四個房市交易資訊以進行關係的實證測試。結果顯示Facebook與PTT社群總情緒聲量或區分為正面、負面、中立情緒之測試，可發現對房價與房屋交易量有顯著影響力，但對房屋流通天數與議價空間就只有PTT社群情緒聲量發現影響力。在Granger因果關係檢定方面，房價與交易量與社群情緒聲量互為因果，也就是發生回饋的情況，然社群情緒聲量僅單向影響流通天數，並與議價空間有很弱的因果關係。透過觀察衝擊反應函數，Facebook社群情緒聲量對房價在衝擊發生的初期有正向且顯著的影響，PTT社群情緒聲量則是對房價、議價空間分別有相對較長期間的正向、負向影響。本文證實社群媒體對於房市具有一定程度的影響力，透過觀察社群情緒聲量能對房市變化更有掌握。

關鍵字

社群媒體情緒 ； 房市 ； 網路聲量 ； 臉書 ； 網路論壇

並列摘要

Online social networks have become a popular channel through which individuals obtain opinions and evaluations that enable them to make informed decisions. Studies have determined that traditional media are a key factor influencing various markets. The present study explored the relationship between social media and housing market changes. The OpView database was searched to retrieve online responses posted between 2016 and 2020 on various social media platforms (i.e., Facebook and PTT). Social media sentiment indices, which were constructed on the basis of the collected online responses, were used to analyze the correlations of the responses with housing price, transaction volume, time-on-the-market, and price concessions. The results indicate that the general, positive, negative, and neutral online responses posted on Facebook and PTT significantly influenced both housing prices and transaction volume. However, only the online responses posted on PTT significantly affected time-on-the-market and price concessions. The results of a Granger causality test revealed the two-way causal relationships of housing prices and transaction volume with online responses, indicating the presence of a feedback mechanism. However, online responses only affected time-on-the-market in one direction, and their causal relationship with price concessions was weak. The results obtained using the impact response function indicate that the effect of Facebook online responses on housing prices was only significant and positive during the initial period of shock. By contrast, the positive and negative effects of PTT online responses on housing prices and price concessions, respectively, persisted for a longer period. We determined that online social media responses influenced the housing market. Through an analysis of social media sentiment, we obtained a clearer understanding of housing market fluctuations.

並列關鍵字

Social Media Sentiment ； Housing Market ； Online Response ； Facebook ； Internet Forum

參考文獻

范清益，2010，「買屋賣屋「殺」很大！－議價空間與住宅不動產市場流動性之影響因素分析」，土地問題研究季刊，9：82-91。(Fan, C. Y., 2010, “Intense Competition in Buying and Selling Houses－An Analysis of Factors Influencing Bargaining Space and Market Liquidity in the Residential Real Estate Market”, Land Issues Research Quartely, 9: 82-91.)

殷志偉與劉正，2020，「非核家園的民眾意向：網路輿論的大數據分析」，選舉研究，27：49-92。(Ying, C. W. and J. Liu, 2020, “Public Orientation on the Nuclear-free Homeland Issue: A Big-Data Analysis of Online Comments”, Journal of Electoral Studies, 27: 49-92.)

程輝與劉雲，2008，「基於時間序列的網路輿情預測模型」，網際網路技術學刊，9：429-432。(Cheng, H. and Y. Liu, 2008, “An Online Public Opinion Forecast Model Based on Time Series”, Journal of Internet Technology, 9: 429-432.)

楊立偉，2020，社群大數據：網路聲量、口碑及輿情分析，新北市：前程文化。(Yang, L. W., 2020, Social Big Data: Listening & Analytics, New Taipei City: Future Career Publishing.)

楊茜文、朱芳妮、呂少毫與陳明吉，2022，「房市參與者之關注會影響房價嗎？」，管理與系統，29: 495-523。(Yang, C. W., F. N. Chu, S. H. Lu, and M. C. Chen, 2022, “Will Attention of Market Participants Influence Housing Prices?”, Journal of Management & Systems, 29: 495-523.)