LSI (Latent semantic analysis) is used to find similar documents in a corpus.
The typical use case is as follows:
- Given a document d, find the documents most similar to the given document.
Gensim's typical SVD allows only querying a document against a corpus to find the closest matching documents. The idea is to use LSI to find term similarity.
Python
LSI (Latent semantic analysis) is used to find similar documents in a corpus.
The typical use case is as follows: