Decision tree algorithms such as See5 (or C5) are typicallyused in data mining for classiÂ¯cation and prediction purposes. In thisstudy we propose EXPLORE, a novel decision tree algorithm, which is amodiÂ¯cation of See5. The modiÂ¯cations are made to improve the capabilityof a tree in extracting hidden patterns. JustiÂ¯cation of the proposedmodiÂ¯cations is also presented. We experimentally compare EXPLOREwith some existing algorithms such as See5, REPTree and J48 on severalissues including quality of extracted rules/patterns, simplicity, and classiÂ¯cation accuracy of the trees. Our initial experimental results indicateadvantages of EXPLORE over existing algorithms.