Classification and sentiment analysis and research is one of the most important issues in recent times, especially in the realm of language is Persian. Identify the user sentiments about an issue through a process of expressing a sense of individual and output data analysis work is in the context of Hijab of natural language processing and computational linguistics, artificial intelligence is a subsidiary. The need to categorize and production data sources in our field and that's the dress code, according to the news raised in this context and user reaction to the news texts, to provide a data set with an algorithm that on it has been implemented. The algorithm proposed in this study using text mining techniques and algorithms applied weight to words and try to identify the type of polarization of the comments gathered from official sites built in 1000. Analysis on the words in the comments feature has been conducted. It focused on the importance of the word in a sentence.