各類新聞與正負面情緒對房市之影響：文字探勘之應用

Application of Text Mining: The Influence of Media Sentiment on Real Estate Market By Different News Topics and Positive/Negative Sentiment

郭偉傑(Kuo Wei-Chieh)

指導教授 ： 陳明吉

國立政治大學/商學院/財務管理學系/碩士(2020年)

https://doi.org/10.6814/NCCU202000812

摘要

本研究透過於聯合知識庫蒐集2009年至2018年有關房市、股市、勞動市場與人口統計新聞共計70,533篇，並應用文字探勘與情緒分析技術，利用財金領域辭典作為分析情感的依據，計算各個不同新聞主題每月所隱含的情緒指標，來研究房市參與者會受到哪些主題的新聞所影響進而做出相對應的房市交易決策行為改變房價、房屋交易量、房屋流通天數與房屋議價空間。另外為了分析房市交易資訊是否具有正負面影響力不同的情形，本研究在計算情緒指標上也額外分別建立了正面與負面的情緒指標，來探討房市參與者較容易受到正面亦或是負面情緒所影響；此外，本研究為了探討媒體情緒與房市交易資訊之因果關係，亦採用Granger因果關係檢定來進行驗證。

　　本研究發現，房市媒體情緒將能顯著影響下一期的房價、房屋交易量、流通天數與議價空間，而將房市主題拆分為更細的主題後，也發現如租屋、房屋供給、房市政策與房市信用狀況新聞媒體情緒皆能顯著影響下一期的房價。除了房市以外的市場中，本研究也發現股市、勞動市場、人口統計媒體情緒也會顯著影響下一期房價，是故本研究證實了不只房市新聞媒體情緒將會顯著影響下一期房市交易資訊，若將房市媒體情緒做更細緻的拆分或是納入不同市場的媒體情緒，也能對未來房市交易資訊具有顯著的影響。

　　正負面媒體情緒中，本研究發現許多不同主題新聞中正面情緒與負面情緒影響力不同之現象；因果關係驗證上，本研究發現房市媒體情緒波動會造成房價、房屋交易量、房屋流通天數與房屋議價空間之改變，具有顯著因果關係。

關鍵字

文字探勘 ； 情緒分析 ； 房地產市場

並列摘要

Base on the vigorous development of text mining and sentiment analysis in recent years, it has also been gradually applied in various financial markets. This research collect news about the housing market, stock market, labor market and demographics from 2009 to 2018 via Udndata.com and capture 70,533 articles. Through text mining and sentiment analysis techniques, we constructed a series of monthly sentiment for every news topic and examine the relationship between the media sentiment and the housing market. Besides, we also separately established positive and negative sentiment index to explore whether housing market participants are more susceptible to positive or negative sentiment. In the end, we also used causality test to check the relation between sentiment and the houseing market.

The empirical results shows that the housing market sentiment will significantly affect the trading volume and the wiggle room in the next period. Also, after splitting the housing market media sentiment into more detailed themes, it also found that such as rental, housing supply, housing market policy and the credit situation media sentiment can significantly affect the house prices. In markets other than the housing market, we also found that stock market, labor market, demographic media sentiment will also significantly affect the house prices. To conclude this study, we confirmed that not only the housing market news media sentiment but also stock market, labor market and demographics media sentiment significantly affect the housing market. Besides, we found the Positive/Negative sentiment influence the housing market differently. In the end, we also found house media sentiment would Granger cause the housing market.

並列關鍵字

Text Mining ； Sentiment Analysis ； Real Estate Market

參考文獻

英文參考文獻

Baker, M., Wurgler, J., (2006), Investor sentiment and the cross‐section of stock returns, Journal of Finance, 61, 1645–1680.

Ball-Rokeach, S., DeFleur, M. L., (1976), A Dependency Model of Mass Media Effects. Communication Research, 3(1), 3-21.

Barber, B., & Odean,T., (2008), All That Glitters: The Effect of Attention and News on the Buying Behavior of Individual and Institutional Investors. The Review of Financial Studies, 21(2), 785-818.

Baumeister, R., Bratslavsky, E., Finkenauer, C., Vohs, K. D., (2001), Bad is stronger than good. Review of General Psychology, 5, 323–370.