Abstract
Social media sites provide us facility to get the data which is generated by the customers, this data is available in large amount. Companies not only need to analyze the data generated on social media sites by their own customers, but they also need to analyze the data generated by the competitor's customers. In this research paper, we collected the data of six fast food chains KFC, Pizza Hut, Subway, Dunkin Donuts, Domino's Pizza and McDonald from twitter. We analyze 25,000 English tweets of each company. We used the lexicon for sentiment analysis, for each tweet we compare every word of the tweet with lexicon and determine that either this tweet contains more positive words or negative words. The main concern of this research is to calculate the reputation from the collected tweets in Twitter. To do so, an English sentiment lexicon has been used to classify tweets then beta probability function used to calculate reputation score of every restaurant based on the calculating of number of positive/negative words and tweets. Experimental results show some statistical information such as comparing the number of positive/negative tweets, the occurrence of top 10 positive/negative word, and reputation of every restaurant based on word-level and sentence-level statistics calculation.