Dissertation (MSc. Computer Science)Forecasting of stock market trends has been an area of great interest to researchers who are attempting to uncover the information hidden in the stock market data and to traders who wish to profit by trading stocks. An accurate forecasting of stock market trends may yield profits for investors. Forecasting of stock price trend is regarded as a challenging task. Due to the complexity of stock market data, development of efficient models for forecasting stock market trends is highly challenging. Applications of data mining techniques for stock market forecasting are an area of research which has been receiving a lot of attention recently. This study presents the development and evaluation of a decision tree and naïve Bayes hybrid model for stock market next day’s trend forecast in Dar-Es-Salaam Stock Exchange (DSE). Historical DSE data is used in the present study to extract features that can cause change in stocks price trends. In the developed hybrid model, decision tree is used to select the subsets of relevant features and naïve Bayes is used to produce a stable model for forecasting stock market trends. This study found that, the proposed hybrid model outperforms both the baseline decision tree and naïve Bayes models. It is found that features selection using decision tree employed in this study significantly improved the trend forecasting performance in stock market. It can be concluded from this study that, the decision tree and naïve Bayes hybrid model performs well and is reliable in stock market trend forecasting.