You are looking at a content-based recommendation algorithm. AFAICT StackOverflow looks at the tags and words in the title and finds questions that share some of them. It can be implemented as a search for the nearest neighbor in a space where documents are represented as TF-IDF vectors .
Implemented, go with any Django search engine that supports stems, stop words, non-line matches, and tf-idf scales. The algorithmic complexity is low (only a few index queries), so it doesn't matter if it is written in Python.
If you do not find that the search engine does what you want, leave the source and stop words in the search engine, call the search engine for individual words and make your own tf-idf with an rating that supports similar tags.
source share