A Study On Term Connection Oriented NLP Technique And Its Applications

With the development of social informationization and computer hard/software level, Natural Language Processing (NLP for short) techniques have gradually attracted more attentions for computer applications and artificial intelligence as popular topics, taking human Emulation technique with full capabilities of listening, speaking, reading and writing for example. A new specialty field, Computational Linguistics, came into existance concerning NLP techniques, which is in itself a cross subject study widely supported by Linguistics, Psychology, Psycholinguistics, Brain Science, Computer Science, Phylosophy, Logics, Artificial Intelligence, Mathematics, Information, Literature, Arts ect. Ever since the Mid 1900s, a few achievements have been made in aspects of effective techniques, yet we still have a long way to go. At the beginning of this new century, it's of high nessesity to reevaluate the corresponding NLP research background, purpose and approaches. Further more, creativity is appreciated on our current techniques towards more applicable and future-oriented new tech. It is right in these areas that the present paper makes its own effort, hoping that it will stimulate deeper insights and contribute to NLP technique's further development.Under the National Science Funded Project, Computer Aided Literature Creation Research-Poem, Ci, Ballad, Couplet (60173060, 2002-2004), this paper proposes and primarily constructs a Term Connection Technique for NLP (TCT for short), and further applies it in the Poetry Processing System (PPS for short). Both theratical and practical studies prove that TCT is a highly comprehensive, widely applicable, easily manipulated effective NLP technique inheriting and developing the now existing techniques. The main contents of this paper are as follows:Chapter 1 introcuces TCT technology background with personal insights, which includes NLP fundamental difficulties, its basic aims at present and related technology developing periods and trends. All those difficulties lie in Cognition and Uncertainty features of Natural Language, while its purpose now is certain defined Target-Specific Language Intelligent Cognition Simulating Technique. This periodic goal was proposed on thorough analyses of NLP fundamental difficulties, available computer technology, present Artificial Intelligence and NLP research achievements, thus it is of high feasibility. NLP technology can be divided into three periods: discovery of technology, rise of technique principles, and rise of statistical analyses. Its developing inclinationmight be in favor of sample technique as the main, together with principles and statistics as auxiliary techniques.Chapter 2 brings forward basic principles, technical structure, philosophical foundations, and methodology of TCT in general. The basic principle of TCT is to build up a dynamic corpus based on defined Target-specific Natural Language environment, and then construct sample knowledge corpus and language rule corpus on the previous dynamic corpus, so finally, certain defined Target-specific Natural Language can be processed on these corpuses. TCT here enjoys an integrate, concise and sound technical structure, including TCT Knowledge Representation Technique (TCTR), TCT Knowledge Obtaining Technique (TCTO), TCT Language Analysis Technique (TCTA), TCT Language Evaluation Technique (TCTE), TCT Language Construction Technique (TCTC), TCT Language Modification Technique (TCTM) and TCT Language Input and Output Technique (TCTIO). The philosophical background of TCT is Yi (a Chinese philosophy of changes), which is the most striking feature of this paper. The study of Yi is an antient classical philosophy with its simple materialistic dialectic theories and systematic view. This is sure to powerfully guide our cognition modeling. Besides, as an NLP oriented language theory, Neurological Cognitive Linguistics is the foundamental theory of TCT, which emphasize language theory's accordance with brain nerves and bridges together advantages in Connectionism, Symbolism etc. The primary met...