Preprocessing For Crawler Of Short Message Social Media

摘要

Social media can be utilized source information in the form of text that is widely exploit as an analytical tool to understand the attitudes, preferences and opinions of society. Companies can use for produce decisions about the needs, attitudes, opinions or trends about customers or potential customers. One of the popular social media now is Twitter. Research aims to design web applications crawl on twitter social media Indonesian language for Natural Language Processing needs. Methodology promote in this study is crawl twitter and preprocessing of data that includes parsing and tokenize, emoticon conversion, cleansing, case folding, normalization, negation handling, stop word and stemming. Research can be further exploited in text processing for classification of analytical sentiments.