Design And Implementation Of Heuristic Analogy Translation Mechanism In IHSMTS

Machine translation technology is approved to be an effective way in conquering the language barrier. With the growth of international communication and development of the World Wide Web (WWW), it was more and more seriously emphasized. Machine translation is expected to present high quality and fast translation service. In the past fifty years machine translation research has zigzagged to a practical ground after some breakthroughs. However, the research of machine translation is far from maturity, and some difficulties slow down the race of the research. Firstly, quality of machine translation output is not yet satisfying to meet the need of high quality of translation. Secondly, the bottleneck of automatic knowledge acquisition decreases the extensibility of the machine translation system, on the other hand, the hardship of knowledge maintenance increases when the knowledge is acquired manually.Interactive Hybrid Strategy Machine Translation System (IHSMTS) is a practical MAT platform, which attempts to resolve the difficulties mentioned above. Combining the rule-based machine translation (RBMT) strategy with the example-based machine translation (EBMT) strategy, Case Pattern, a new knowledge representation method, is proposed in IHSMTS, which helps to overcome the difficulty of knowledge acquisition. Enlightened by rule-based language analysis, an intelligent analogy translation mechanism is proposed in IHSMTS, which improves the quality of translation.To realize the intelligent analogy translation, the Heuristic Analogy Translation Mechanism (HATM) is proposed in this thesis, which is...