CROSS LANGUAGE INFORMATION RETRIEVAL UNTUK DOKUMEN BAHASA INDONESIA DAN BAHASA MANDARIN

Abstrak

We develop a CrossT LanguageF Information RetrievalE (CLIR) application for Indonesian and Chinese language documents. The main purpose of this application is to find documents that are relevant to document input. We applied Parallel Corpus for build in dictionary. This system makes 500 words . We applied Vector Space Model (VSM) for searching algorithm. VSM measure relevancy for each document in the form of weight. System accurately selects relevant documents with average accuracy of precision and recall for Chinese query are 0.857479 and 0.030692, for Indonesian query are 0.86473 and 0.0324.  The application used 50 queries in Indonesian and Chinese language.

Keywords

Parallel Corpus, Vector Space Model, Indonesian, Chinese.