Question

Briefly explain what LSI (Latent Semantic Indexing) achieves that the other retrieval models cannot.

(( data mining question ))

Answer 1

Compared to other retrieval models LSI gives more performance as it uses Singular Value Decomposition(SVD) compared to plain vector space model.

it is faster, compared to other dimensionality reduction models Since it only involves decomposing term document matrix.

By the usage of reduced representation in LSI it is able to remove some noise from the unstructured data which will enhance its performance.

The vector space need only be constructed once for the entire collection which would mean that the further retrieval of the results would not take much time.