Paper Info
Title
A Comparative Study Of Topic Identification On Newspaper And E-Mail
Abstract
This paper presents several statistical methods for topic identification on two kinds of textual data: newspaper articles and e-mails. Five methods are tested on these two corpora: topic unigrams, cache model, TFIDF classier, topic perplexity, and weighted model. Our work aims to study these methods by confronting them to very different data. This study is very fruitful for our research. Statistical topic identification methods depend not only on a corpus, but also on its type. One of the methods achieves a topic identification of 80% on a general newspaper corpus but does not exceed 30% on e-mail corpus. Another method gives the best result on e-mails, but has not the same behavior on a newspaper corpus. Me also show in this paper that almost all our methods achieve good results in retrieving the first two manually annotated labels.
Year
DOI
Venue
2001
10.1109/SPIRE.2001.989770
EIGHTH SYMPOSIUM ON STRING PROCESSING AND INFORMATION RETRIEVAL, PROCEEDINGS
Keywords
Field
DocType
routing,statistical analysis,information retrieval,language model,testing,natural languages,speech recognition
tf–idf,Information retrieval,Cache,Computer science,Newspaper,Natural language,Artificial intelligence,Natural language processing,Text categorization,Vocabulary,Language model,Statistical analysis
Conference
Citations 
PageRank 
References 
8
0.61
5
Authors
5
Name
Order
Citations
PageRank
Brigitte Bigi133627.76
Armelle Brun213821.49
Jean-Paul Haton338065.42
Kamel Smaïli412025.18
Imed Zitouni561246.39