Investment portfolio optimization based on risk and trust management

The goal of this paper is to develop a system able to coordinate a trader in optimizing a stock market portfolio in order to improve the profitability of a short or medium time period investment. The system is able to classify the risk and quantifies its effect on an investment based on sentiment analysis, certain characteristics and the traders level of confidence. It also uses an ontology to describe the markets status and expectations based on future events (text features) and to determine the relationship and correlation between news articles. The system also verifies the effect of a similar percentage of positive and negative news articles, and which of this information will influence more the stock price movement and trend. We proposed a multi-agent system that uses text mining, ontology, sentiment analysis, trust and risk models in order to choose the appropriate mix of investments to minimize the risk and maximize the gain on a stock portfolio. A prototype was developed on which we validated our research.