Consider the following document collection D = {D1, D2, D3} (given as one document per line):

D1 => Silly Sally Sleepy Sally

D2 => Seven Silly Sheep

D3 => Silly Sheep Should Sleep Silly

Assume that the stopword list contains the word Should, and words are stemmed (that is, converted to their root).

Show the dictionary and the postings list including all the relevant statistics computed, such as raw tf-idf values shown explicitly as (tf,idf) with each document in the postings list), for implementing (uncompressed) inverted index structure for Vector Space Ranked Retrieval in an easy-to-read format. Assume that raw term frequency factor is the count of the number of term occurrences in a document (rather than the normalized, log-dampened value) and the inverse document frequency factor is the reciprocal of the fraction of documents that contain the term (rather than its logarithm).

What are the relevance scores and the ranking of the documents for the query: Silly?

Does the ranking change if we define term frequency factor as the normalized fraction of the term occurrences in a document (rather than the raw count).