'Graph Theory' Category
Discussion of Section 3 of TextRank
Section 3 of TextRank: Bringing Order Into Texts covers the first application of the TextRank approach proposed in section 2. The authors have chosen keyword extraction to demonstrate the capabilities of the approach. Keyword extraction is the problem of determining the keywords that best describe a document. It can be thought of as a precursor [...]
Popularity: 6% [?]
Discussion of TextRank sections 1 and 2
The first two sections of TextRank: Bringing Order into Text by Mihalcea and Tarau provide valuable background information and introduction to applying graph theory (see http://en.wikipedia.org/wiki/Graph_theory for background) to NLP based programs. As we saw in Brin and Page last week, we can use graph based approaches for determining node importance through citation analysis. [...]
Popularity: 3% [?]
TextRank by Rada Mihalcea and Paul Tarau
This week we will be continuing our graph-theory based approach to NLP and take a look at TextRank: Bringing Order to Texts. The paper claims to show us how to use graph-ranking approaches in some unsupervised learning tasks such as keyword and sentence extraction.
TextRank, NLP, natural language processing, unsupervised learning, PageRank, HITS, computer science, algorithms
Popularity: [...]Popularity: 5% [?]
The Anatomy of a Search Engine
The Anatomy of a Search Engine is the Paper of the Week for the week of January 22, 2007.
OK, I’ll admit I’ve read this one before (two or three times, actually) but it’s going to serve as the intro to several other pieces on using Graph Theory for doing NLP. Plus, it is the [...]Popularity: 6% [?]
