I have come across this topic of Latent Semantic Analysis. I could not understand the concept of Latent Semantic Analysis.
1 - Can Latent Semantic Analysis be used by humans in the clustering of the data-sets ? For convenience let us consider the data-sets to be a two dimensional sets. Can the humans cluster the data-sets manually without the help of clustering algorithms?
2 - Are the clustering algorithms using the concept of Latent Semantic Analysis?How?
Latent semantic indexing is a technique that relies on SVD decomosition of term-document matrix. It enables semantic information retrieval by maping similar documents closer regardless of terms used in the text. It has potential to overcome problems of synonyms in information retrieval.
Clustering can be connected to latent semantic indexing in various ways. For example, as latent semantic indexing is technique that works well for homogenious collections of documents, it is usefull to cluster document collection before application od LSI and apply LSI on separate clusters. This paper can help:
http://www.sciencedirect.com/science/article/pii/S0306457304001384
Also, there is clustering algorithm that relies on SVD (as LSI):
http://www.cs.put.poznan.pl/dweiss/site/publications/download/iipwm-osinski-weiss-stefanowski-2004-lingo.pdf
- Badges
- Science topic
- Similar topics
- Computer Science
- Data Mining