NLP And ML Based Text Classification And Its Applications

This dissertation mainly dicussed the Natural Language Processing (NLP) and Machine Learning (ML) based text classification, in which a new mean-variance based feature reduction was raised and anlyized. The relation between the classification effectivness and the dimension of feature space was examined, together with using two different learning methods in the second stage. Also, the NLP and ML were described including both the theoretical, algorithm aspect and the engineering aspect, in which an efficient data structure was given. The application of text classification in search engine and information filtering was discussed. Furthermore, the exact effectiveness and efficiency the TC can help the search engine gain and the application in personalized information push in an exact system of recruiting information push were discussed. The possible improvement was also given in this dissertation.The text classification can be achieved in two stages, using NLP and ML correspondingly. Thus, evidently, TC has its theoretical value in pushing the research of NLP and ML. Its applications in search engine by improving its effectiveness and efficiency, in the personalized information push service, in the pattern of information gaining and content safety are also very meaningful. So, text classification has become an important task both in theory and engineering.