Regression-Based Microblogging Influence Detection Framework for Stock Market

Abstract

Microblogs and social networks have become a valuable resource for mining sentiments in various fields. The sentiments posted on the web have reportedly influenced the trading and investment decisions and activities taking place in the stock exchanges. In this study, we have investigated explored the effects of microblogs on Chinese Stock Market. We have particularly focused on whether measurements of collective mood states (sentiments and persuasions) derived from large-scale microblogging posts are correlated to the values of Chinese stock market over a period of time. We have proposed a Regression-Based Microblogging Influence Detection Scheme (RMIDS) as a framework to detect the influence of microblogging posts on stock market in this paper. Our results showed that sentiments in microblogging has significant influence on the market, while the persuasion posts trying to interfere the stock market, by taking the advantage of lack of regulations in microblogging, do not succeed.

Using social media mining technology to improve stock price forecast accuracy

“…Behavioral finance has proven that financial decisions are clearly driven by emotions (Rubbaniy, Asmerom, Rizvi, & Naqvi, 2014). The sentiments of reviews posted on social media reportedly affect trading, investment decisions, and activities in the stock exchanges, and thus become a valuable resource (Zhu et al, 2014). In recent years, significant progress has been made in sentiment tracking techniques that directly extract public sentiments from social media content.…”

“…With the rise of social media and the promotion of relevant platforms for user-generated content, user opinions have begun to play a greater role in the stock market. Zhu et al (2014) reported in 2008 that about 25% of adults are indirectly dependent on investment advice spread through social media. In response to the increase in online news and social media, web-mining-based forecasting methods have been extensively studied to improve the performance of financial market forecasts (Xu, Li, Jiang, & Cheng, 2012).…”

“…Kim, Jeong, and Ghani (2014) designed a method that included natural language processing classification and the extraction of sentiments and opinions expressed by authors, but the effects were not satisfactory. Zhu et al (2014) proposed a Microblogging Influence Detection Scheme (RMIDS) to detect the impact of microblogging posts on the stock market. Their results showed that the sentiments in microblogging is significantly related to Chinese stock prices.…”

“…Zhu et al () proposed a Microblogging Influence Detection Scheme (RMIDS) to detect the impact of microblogging posts on the stock market. Their results showed that the sentiments in microblogging is significantly related to Chinese stock prices.…”