Latent semantic indexing (LSI) is a completely automatic yet intelligent indexing method,widely applicable,and a promising way to improve user' s access of many kinds of textual materials. LSI tries to overcome the problem of lexical matching by using statistically derived conceptual indices instead of individual words for retrieval.In this thesis LSI is introduced into case retrieval,and the application of LSI represented also. In order to preprocess the text of aircraft fault,a new automatic indexing method based on n-gram is proposed. Large quantities of experiments to adjust the parameters of LSI model are done so that the performance of case retrieval system is excellent.An LSI-based case retrieval system is developed and some work is done to optimize the system. |