Leyton is an international consulting firm that helps businesses leverage financial incentives to accelerate their growth and achieve long-lasting performance.
Stock market prediction is a major economic task in trading activity planning. It has also presented a crucial topic in several research fields such as engineering, mathematics, finance and computer science.
Social media can improve the accuracy of the stock market
Predicting the stock market is a task proved to be challenging due to the complex nature of the stock market. Indeed, its strong dependence on various factors such as unpredictable political events, the launch of new products or the general mood of the society based on the opinions and emotions of the general public [1] on one hand. And on the other hand, due to the complexity of the dynamic fluctuation of the stock market which can be modeled from the numerical/textual data.
This modeling makes it possible to provide quantitative and qualitative information on the movement of stock market prices: quarterly and annual financial reports, press or web articles. As a result, stock market prediction has improved as the number of such dynamic factors exploited in stock market analysis is large [2].
With the huge emergence of social media, the abundance of information about the opinions and feelings of the general public has become a very interesting key that can be incorporated to analyze the behavior of the public towards a certain product and then towards its investment. In addition, exploiting the social mood provided by social media can also improve the accuracy of stock market forecasting models, since it is one of the most important factors influencing the price of its action [3].
Opinion analysis from social media
Recently, social media has received a lot of attention from companies in order to study its possible exploitation in the process of predicting the movement of stock prices. Social media in general and social networking in particular allow people to share their innermost thoughts or express their most innocent feelings/opinions about real world events [4].
Twitter, for example, is the richest body of data available for free and quickly updated, providing businesses with real-time alerts on daily stock price movements. Also, concise insights into future consumer buying behavior. All in all, the exploitation of textual data extracted from social media in addition to numerical data is proved to be very useful in increasing the accuracy and quality of stock market forecasts.
One of the main areas of innovation today is the real-time monitoring of public sentiment and opinions on market trends: by the analysis of opinions from social media. Due to its extreme importance, the application of this axis in finance has been the goal of several machine learning projects [5]. Indeed, the application of text mining models and the analysis of opinions on social networks offers an excellent way to predict future fluctuations in the stock market through the public speculations.
Recently, many research and developmentprojects have addressed the problem of being able to correlate the activity of this data with fluctuations in stock prices. And highlighted the usefulness of social networks as a real-time source of data that must be processed and interpreted in order to predict stock market fluctuations.
L. Hung, the Presidential Election and the Stock Market in Taiwan. Journal of Business and Policy Research.
E.F. Fama, The behavior of stock-market prices, The Journal of Business.
X. Zhang, H. Fuehres and P. A. Gloor, Predicting stock market indicators through twitter “I hope it is not as bad as I fear”. Procedia-Social and Behavioral Sciences.
A. Cohen, Unsupervised gene/protein named entity normalization using automatically extracted dictionaries.
M. G. Berry and G. S. Linoff, Data mining techniques: for marketing, sales, and customer relationship management
