likarajo / moviesearch

Search for movies based on input words by using TF-IDF and Cosine Similarity on the movie summaries

Information Retrieval

Term Weighting

• Local: How important is the term in this document? => Term Frequency (TF)
• Global: How important is the term in the collection? => Document frequency (DF)

TF-IDF:

• Terms that appear often in a document should get high weights : TF
• Terms that appear in many documents should get low weights: IDF

w_i,j = weight assigned to term i in document j
tf_i,j = number of occurrence of term i in document j
N = number of documents in entire collection
n_i = number of documents with term i

w_i,j = tf_i,j log( N / n_i )