Prediction of attractive evaluation objects based on trend rules and topic dictionary

This paper proposes an improvement method for prediction of attractive evaluation objects based on trend rules. The trend rules represent relationships among evaluation objects, key phrases, and numerical changes related to the evaluation objects. The trend rules are inductively acquired from text sequential data and numerical sequential data. The method assigns evaluation objects to the text one by activating topic dictionary. The dictionary stores key phrases representing the numerical change. It can expand the amount of the learning data. It is anticipated that the expansion leads to acquire more valid trend rules. This paper applies the method to a task which predicts attractive stock brands based on both news headlines and stock price sequences. It shows that the method can improve the detection performance of evaluation objects through experiments.