An Intelligent Business Advisor System for Stock Investment

Paper 1:  An Intelligent Business Advisor System for stock investment

The paper describes an intelligent system that assists small stock investors in determining the trend of the stock market , and makes recommendations for investment for its users based upon the past patterns and market conditions. The topic is genuine since the stock market of Honk Kong is a huge and complex one with millions of transactions occurring daily. New investors, when trying to invest are confused by the complexity of the financial market and its unpredictiveness. The mentioned system tries to ease the burden of the small investors by carrying out a lot of technical analysis as well as providing them with recommendations for future actions based on the same. The paper focuses on the rationale behind stock and technical analysis, and then explains the architecture of the system designed, followed by its application and test results. The system designed in the paper claims to be an expert system but it goes beyond that and would more rightly be defined as a Suggestion Model as per Alters’ Taxonomy whose intended users are small stock investors .For novice users , it provides tutoring features and supports analysis of the study behind the rationale for the investment. It also includes features that allow skilled users to modify the theories and rules as per their need and obtain predictions for those. Pattern recognition of trend signals and the usage of heuristics are used by the system to provide recommendations for the end user.  The StockAdvisor system was developed using KAPPA tool which allows it to work on MS windows platform. For the case study it was deployed on Microsoft Excel 5, hence providing its users with a familiar and easy to use system. The system makes use of stock market features like Trendline, Momentum and Moving Average further divided into sections for the final recommendations using two level dependencies between Kappa in the background for calculations and Excel in the foreground for results display. The prototype had 33 stocks in its database and when tested for investment within a period of a year the net result by its recommendations was 22% return which far exceeded the inflation rate of 10%. Although the system provides a great inexpensive and friendly for small investors to assist in decision making, there is a room for further improvement as mentioned in the paper. The tool can be upgraded to include Fundamental Analysis along with technical analysis. It can also make use of case-based systems for further rule improvements.