Machine Learning for Mobile

TF-IDF

The TF IDF formula gives the relative importance of a term in a corpus (list of documents), given by the following formula: 

Where:

• tf_i,j = number of occurence of i in j
• df_i = number of documents containing i
• N = total number of document

Consider a document that contains 1,000 words, wherein the word rat appears 3 times. The term frequency (TF) for rat is then (3/1000=) 0.003. Now, in 10,000 documents, the word cat appears in 1,000 of them. Therefore, the inverse document frequency (IDF) is calculated as log(10000/1000) = 1. Thus, the TF-IDF weight is the product of these quantities is 0.003 * 1 = 0.12.

The words or features in the ...