An Information Retrieval Algorithm to Extract Influential Factors

Past literatures showed that there are many factors that can be used to assess company’s performance but only a limited number of factors are needed to efficiently assess its performance. The aim of the study is to develop an algorithm that can extract a minimum set of factors that can be used to assess companies’ performances. Stock price was used as the dependent factor. The factors extracted are known as influential factors because these factors were found to have strong influence on the stock price. The objectives of the study were to obtain a comprehensive influential factors from past literatures, develop
an extraction algorithm that can identify influencial factors, and present factors that influenced companies’ stock prices. Data consisted of financial factors that were
obtained from financial documents of distressed companies and non-distressed companies listed on a stock exchange. The extraction algorithm was developed and implemented using Matlab programming language. Results showed that out of 33 factors, 5 factors were found to be the minimum set needed to assess the companies’ performances. These were debt, investment, total asset, asset turnover, and working
capital. The algorithm were tested on other dataset and results produced more than 70 percent of positive feedback. This indicates that the algorithm was able to produce a good
model. The extraction algorithm developed showed that influencial factors produced could be used as guideline for companies to monitor and strategize ways for business
improvement.