Research on the Fast Retrieval Algorithm of English Sentences Based on Simhash

In the current situation, the case-based translation method is becoming more and more popular. The advantage of using machine translation is that machine translation does not need to analyze a sentence word by word, but to replace it under the same word meaning in the original language database. When there are more cases, the translation results will be more accurate. This paper mainly studies the case-based machine translation method. In order to find more accurate similarity, some similar cases of the system in the retrieval process. This paper first introduces the current situation of case-based machine translation and the practical problems of the system, and studies the relevant principles of the simhash algorithm and the principles of the simhash algorithm, so as to realize the purpose of the simhash algorithm to quickly retrieve similar cases, and carefully handle the details in the process of using the simhash algorithm, which can reduce the number of similar cases to be retrieved. With the time, the larger the corpus size, the more accurate the results can be brought to the retrieval in the detection. If synonyms appear in the similar situation, the similarity of the two words can be calculated by the above method, and the real similarity can be seen more objectively from the results.