An automated stock recommendation system from stock investment research using domain specific information extraction
Tayida Tapjinda, Potsawee Vechpanich, Nutchaya Leelasupakul, Nakornthip Prompoon, Chate Patanothai · 2015
The rise of internet web-based application and smart phone ease access to the stock market, attracting newcomer investors. Most investors have made their decision based on information from stock investment researches published by the brokers. An automated stock recommendation system from stock investment research is introduced in this paper. The system collects several investment researches from multiple broker sites daily, converts the researches from pdf to text file, and extracts stock recommendation from the text and saves them to the database. A web application were also developed to serve the user as an interface to access those extracted information. The developed system is capable of extracting 79% of recommendations with the precision of 85%. The web application function also meets the expected both functional requirements.