Su Mon Khine and Yadana Thein, University of Computer Studies - Yangon, Myanmar

Nowadays web pages are implemented in various kinds of languages on Web and web crawlers are important for search engine. Language specific crawlers are crawlers that traverse and collect the relative web pages using the successive URls of web page. There is very little research area in crawling for Myanmar Language web sites. Most of the language specific crawlers are based on n-gram character sequences which require training documents, the proposed crawler differ from those crawlers. The proposed system focused on only part of crawler to search and retrieve Myanmar web pages for Myanmar Language search engine. The proposed crawler detects the Myanmar character and rule based syllable threshold is used to judgment the relevant of the pages. According to experimental results, the proposed crawler has better performance, achieves successful accuracy and storage space for search engines are lesser since it only crawls the relevant documents for Myanmar web sites.

Language specific crawler, Myanmar Language, rule base syllable segmentation.