Topic-Specific Scoring of Documents for Relevant Retrieval
In: ICML 2005 Workshop 4: Learning in Web Search, 7 August 2005, Bonn, Germany.
There has been mixed success in applying semantic
component analysis (LSA, PLSA, discrete PCA, etc.)
to information retrieval.
Here we combine topic-specific link analysis with discrete PCA
(a semantic component method) to develop a topic relevancy score
for information retrieval that is used in post-filtering
documents retrieved via regular Tf.Idf methods.
When combined with a novel and intuitive
``topic by example'' interface, this allows a user-friendly manner
to include topic relevance into search.
To evaluate the resultant
topic and link based scoring,
a demonstration has been built using the Wikipedia, the public domain
encyclopedia on the web.