The Internet is a large web of documents linked to each other. We can view it as a document graph in which each node corresponds to a document. You will expect documents to link to similar documents; however, web pages sometimes link to other unrelated web pages. This can be by mistake or on purpose, for instance in the context of advertising or attempts to improve search engine rankings. A more trustworthy source such as Wikipedia will probably yield a better graph. However, some Wikipedia pages are very basic stubs, so we may be missing out on quality links.
The cosine similarity is a common distance metric to measure the similarity of two documents. For this metric, we need to compute the inner product of two feature vectors. The cosine similarity of vectors corresponds to the cosine of the angle between vectors, hence the name. The cosine similarity is given by the following equation:
The feature vectors in this recipe are the TF-IDF scores...