Grant: Discovery Grants Program - Individual
University of Ottawa
$23,000.00
discovering subjectivity in language
EF 2007-08
Résumé
The abundance of texts on the Internet makes exciting research projects in computational linguistics possible. Texts need to be interpreted intelligently and affordably, but sometimes more costly processing leads to linguistically more satisfying results. Certain properties of texts require deep processing, even if they can to some extent be discovered with simpler methods. A case in point is sentiment analysis, a new and active area of research. The results have been the strongest in recognizing polarity (e.g., positive/negative movie reviews); sometimes a neutral case is included (e.g., buy stock, sell stock, do nothing). The intensity of opinion has been studied too. All these phenomena are instances of subjectivity in language. In recognizing subjective expressions or classifying texts by opinion values, researchers often apply supervised and slightly primed unsupervised machine learning. The former depends on the availability of annotated lexical resources, whose necessarily manual creation is labour-intensive and obviously subjective.My research has dealt, and will continue to deal, with subjectivity. The language of electronic negotiations contains elements that can predict success or failure, such as expressions of the negotiator's attitude. Short story summarization is a challenging new variation of my work on automatic summarization of news items. Fiction, with its subjective nature, requires methods different than news: complex syntactic and semantic information, including indicators of subjectivity. All my work relies on statistical language processing and on symbolic techniques which include full parsing and linguistically motivated heuristics; this dual strategy will be kept in the proposed research. Work on the language of negotiation will continue once new relevant text data become available. Short story summarization will go on, with much more emphasis on subjectivity. Summarization of news articles will evolve in a novel direction: constructing summaries of multiple documents with varying viewpoints and opinions. Work has also begun on recognizing certain types of opinions in blogs.
Programme
Discovery Grants Program - Individual
Bénéficiaire
Emplacement
Ontario, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data
Tel que publié
Application id
8193-2007
Name nom
Szpakowicz, Stan
Department département
Information Technology and Engineering, School of
Organization id
28
Institution établissement
University of Ottawa
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2007
Competition year année de concours
2007
Award amount
23000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
Discovery grants programs
Group fr
Programmes de subventions à la découverte
Committee code
331
Committee name en
Computing and Information Sciences - B
Committee name fr
Informatique - B
Area of application code
802
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Information systems and technology
Area of application fr
Systèmes d'information et technologie de l'information
Research subject code
2801
Research subject group en
Artificial intelligence
Research subject group fr
Intelligence artificielle
Research subject en
Natural language and speech understanding
Research subject fr
Langage naturel et reconnaissance de la parole
Partie
1
Nb partie
5
Application title
discovering subjectivity in language
Keyword
NA
Application summary
The abundance of texts on the Internet makes exciting research projects in computational linguistics possible. Texts need to be interpreted intelligently and affordably, but sometimes more costly processing leads to linguistically more satisfying results. Certain properties of texts require deep processing, even if they can to some extent be discovered with simpler methods. A case in point is sentiment analysis, a new and active area of research. The results have been the strongest in recognizing polarity (e.g., positive/negative movie reviews); sometimes a neutral case is included (e.g., buy stock, sell stock, do nothing). The intensity of opinion has been studied too. All these phenomena are instances of subjectivity in language. In recognizing subjective expressions or classifying texts by opinion values, researchers often apply supervised and slightly primed unsupervised machine learning. The former depends on the availability of annotated lexical resources, whose necessarily manual creation is labour-intensive and obviously subjective.My research has dealt, and will continue to deal, with subjectivity. The language of electronic negotiations contains elements that can predict success or failure, such as expressions of the negotiator's attitude. Short story summarization is a challenging new variation of my work on automatic summarization of news items. Fiction, with its subjective nature, requires methods different than news: complex syntactic and semantic information, including indicators of subjectivity. All my work relies on statistical language processing and on symbolic techniques which include full parsing and linguistically motivated heuristics; this dual strategy will be kept in the proposed research. Work on the language of negotiation will continue once new relevant text data become available. Short story summarization will go on, with much more emphasis on subjectivity. Summarization of news articles will evolve in a novel direction: constructing summaries of multiple documents with varying viewpoints and opinions. Work has also begun on recognizing certain types of opinions in blogs.