Output details
11 - Computer Science and Informatics
University of Edinburgh
Output title
Streaming first story detection with application to Twitter
Type
E - Conference contribution
DOI
-
Name of conference/published proceedings
Human Language Technologies: The 2010 Annual Conference of the North American Chapter of the Association for Computational Linguistics (HLT '10)
Volume number
-
Issue number
-
First page of article
181
ISSN of proceedings
-
Year of publication
2010
Number of additional authors
2
Additional information
<17> Originality: The paper presents the first constant-space and constant-time approach to solving the First-Story Detection (FSD) problem in the area of Topic Detection and Tracking.
Significance: The paper paves the way for accurate means of detecting novel events in truly massive streams of data, such as Twitter. Previous approaches to the FSD task required storing and querying the entire history of the stream, which makes them impractical on large datasets. The paper overcomes the limitation without a loss of accuracy.
Rigour: The paper was published in an international conference with a 30% acceptance rate.
Interdisciplinary
-
Cross-referral requested
-
Research group
D - Institute for Language, Cognition & Computation
Citation count
32
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-
