'clustering' Category
POTW 5/21/07: Discussion of “A Study on Retrospective and On-Line Event Detection” by Yang, Pierce and Carbonell
Yang’s paper on on-line event detection (”A Study on Retrospective and On-Line Event Detection“) discusses the use of common text retrieval techniques to automatically detect events in news streams.
Imagine that you are responsible for monitoring all the major news feeds in every single country your company does business in order to advise the CEO on [...]Popularity: 13% [?]
POTW 5/21/07: “A Study on Retrospective and On-Line Event Detection” by Yang, Pierce and Carbonell
Paper of the Week for May 20, 2007 is “A Study on Retrospective and On-Line Event Detection” by Yiming Yang, Tom Pierce and Jaime Carbonell.
Popularity: 13% [?]Popularity: 13% [?]
POTW 4/8/07: “Scatter/Gather: A Cluster-based Approach to Browsing Large Document Collections” by Cutting, et.al
This week’s paper is Scatter/Gather: A Cluster-based Approach to Browsing Large Document Collections by Cutting, et. al. This paper, from 1992, takes clustering out of the realm of search, as it was previously used, albeit indifferently, and proposes to use it in a document browsing scenario. In doing so, the authors propose a [...]
Popularity: 4% [?]
POTW: 4/8/07: “Scatter/Gather: A Cluster-based Approach to Browsing Large Document Collections” by Cutting, et.al
This week’s paper is “Scatter/Gather: A Cluster-based Approach to Browsing Large Document Collections” by Cutting, Karger, Pedersen and Tukey. This is one of Doug Cutting’s older works on clustering, pre Lucene fame.
Popularity: 4% [?]Popularity: 4% [?]
POTW 3/26/07: “Lingo: Search Results Clustering Algorithm Based on Singular Value Decomposition” by Osinski, Stefanowski, and Weiss
Finally, a chance to finish up last week’s review on “Lingo” by Osinski, et. al. I first came across Lingo via the Carrot Search and it’s associated Carrot clustering engine. Mr. Weiss also chimes in on the Lucene mailing list from time to time when people ask about clustering Lucene results.
To start, the paper explains [...]Popularity: 4% [?]
