Application of Modified Genetic Algorithm in Feature Extraction of the Unstructured Data

Since unstructured data's quantity is huge and the form is not disunity, thus the extent mining algorithms are hard to mine them. This paper supposes a modified genetic algorithm which extracts the feature from the unstructured data in a high efficiency and makes the further mining conveniently. First of all, it discusses the characteristic of the unstructured data, and then introduces the preprocess of it such as word segmentation, establishing the stop words table and feature extraction. Furthermore, the modified genetic algorithm's operation such as selection, crossover as well as mutation is presented. Finally, the test of result of this modified genetic algorithm is shown, and the test result has proven that the algorithm is effective.