It was just an attempt to automate the process of SUBJECTIVE ANSWER EVALUATION which includes one word,short answers,essays etc. It goes through PREPROCESSING-CASE NORMALIZATION,STOPWORDS are removed to obtain important terms and KEYWORDS in the answer,TOKENIZATION and POS TAGGING is done using NLTK (NATURAL LANGUAGE TOOLKIT). Consequently a table is created and TERM FREQUENCY, INVERSE DOCUMENT FREQUENCY values calculated. The students answer is compared to each correct answer present in the answer database and COSINE SIMILARITY is calculated based on the words used in a similar context. The words used by the student may not be the exact words in the answers database hence SYNONYMS which are SEMANTICALLY SIMILAR are considered by using PYDICTIONARY. The resulting WEB APPLICATION made using DJANGO as back end and BOOTSTRAP as front end .