کلان داده: یادگیری عمیق برای تحلیل تمایلات مالی Big Data: Deep Learning for fnancial sentiment analysis

Introduction Te Internet, as a global system of interconnection, provides a link between billions of devices and people around the world. Te rapid development of social networks causes the tremendous growth of users and digital content [1]. It opens opportunities for people with various skills and knowledge to share their experiences and wisdom with each other. Tere are many websites like Yelp, Wikipedia, Flickr, etc. that use the power of the Internet to help their users make optimal decisions. Furthermore, there are websites that give users the ability to consult with professionals, and one topic that is always popular is investment. Companies like Goldman Sachs and Lehman Brothers have more than 150 years of investment advice. In the Internet age, independent analysts and retail investors around the world can collaborate with each other through the web. Seeking Alpha and StockTwits are two examples of common fnancial social media platforms focused on the stock market, giving their users a way to connect with information and each other and grow their investments [2]. Financial social media brings people, companies, and organizations together so that they can generate ideas and share information with others. It is this media that provides a huge amount of unstructured data (Big Data) that can be integrated into the decisionmaking process. Such a Big Data can be considered as a great source of real-time estimation because of its high frequency of creation and low-cost acquisition. Sentiment analysis (SA) is a common method which is increasingly used to assess the feelings of social media users towards a subject. Te most popular approach performing sentiment analysis is using data mining. Our central idea is to adopt Deep Learning to determine investors’ expectations about the price of stocks and the overall market based on their messages. Te reason why we select Deep Learning methodology rather than data mining is that in data mining, identifying features and selecting the best of those features is the most challenging task to undertake especially in a Big Data.