Investment recommendation by discovering high-quality opinions in investor based social networks

Investor based social networks, such as StockTwist, are gaining increasing popularity. These sites allow users to post their investment opinions in the form of microblogs. Given the growth of the posted data, a significant and challenging research problem is how to utilize the personal wisdom and different viewpoints in these opinions to help investment. A typical way is to aggregate sentiments related to stocks and generates buy or hold recommendations for stocks obtaining favorable votes while suggesting sell or short actions for stocks with negative votes. However, considering the fact that there always exist unreasonable or misleading posts, sentiment aggregation should be improved to be robust to noise. In our work, we study how to estimate qualities of investment opinions in investor based social networks. To predict the quality of an investment opinion, we use multiple categories of factors generated from the author information, opinion content and the characteristics of stocks to which the opinion refers. With predicted qualities of investment opinions, we perform two types of investment recommendation. The first is recommending high-quality opinions to users and the second is recommending portfolios generated by sentiment aggregation in a quality-sensitive manner. Experimental results on real datasets demonstrate the effectiveness of our work in recommending high-quality investment opinions and profitable portfolios.