AUTOMATIC RETRIEVAL AND THE FORMALIZATION OF MULTI WORDS EXPRESSIONS WITH F-WORDS IN THE CORPUS OF CONTEMPORARY AMERICAN ENGLISH

The research problems in this research are 1) how lexicogrammar takes role in determining polarity of F-Word1 and 2) how to formalize it for corpus processing. The data is obtained from the Contemporary American English Corpus (COCA). In this corpus, F-word is proven to be highest in frequency as compared to its distribution across corpora. Corpus methodology is applied by sending queries to retrieve F-Words to COCA interface. Tokens combination surrounding F-words resulted in the phrase and clause unit accompanying F-words, which are significant cues to determine F-word polarity. The polarity is later proven to be not necessarily negative. I also designed a computational resource to allow the retrieval of F-words offline so that users might apply it to any digital text collections