UNT Scholarly Works - Browse

ABOUT BROWSE FEED

Using Encyclopedic Knowledge for Automatic Topic Identification

Description: This paper presents a method for automatic topic identification using an encyclopedic graph derived from Wikipedia. The system is found to exceed the performance of previously proposed machine learning algorithms for topic identification, with an annotation consistency comparable to human annotations.
Date: May 2009
Creator: Coursey, Kino High; Mihalcea, Rada, 1974- & Moen, William E.
Partner: UNT College of Engineering

Topic Identification Using Wikipedia Graph Centrality

Description: This paper presents a method for automatic topic identification using a graph-centrality algorithm applied to an encyclopedic graph derived from Wikipedia. When tested on a data set with manually assigned topics, the system is found to significantly improve over a simpler baseline that does not make use of the external encyclopedic knowledge.
Date: May 2009
Creator: Coursey, Kino High & Mihalcea, Rada, 1974-
Partner: UNT College of Engineering

Text Mining for Automatic Image Tagging

Description: This paper introduces several extractive approaches for automatic image tagging, relying exclusively on information mined from texts. Through evaluations on two datasets, the authors show that their methods exceed competitive baselines by a large margin, and compare favorably with the state-of-the-art that uses both textual and image features.
Date: August 2010
Creator: Leong, Chee Wee; Mihalcea, Rada, 1974- & Hassan, Samer
Partner: UNT College of Engineering

Quantifying the Limits and Success of Extractive Summarization Systems Across Domains

Description: This paper analyzes the topic identification stage of single-document automatic text summarization across four different domains, consisting of newswire, literary, scientific and legal documents.
Date: June 2010
Creator: Ceylan, Hakan; Mihalcea, Rada, 1974-; Ozertem, Umut; Lloret, Elena & Palomar, Manuel
Partner: UNT College of Engineering

Amazon Mechanical Turk for Subjectivity Word Sense Disambiguation

Description: In this paper, the authors discuss research on whether they can use Mechanical Turk (MTurk) to acquire goo annotations with respect to gold-standard data, whether they can filter out low-quality workers (spammers), and whether there is a learning effect associated with repeatedly completing the same kind of task.
Date: June 2010
Creator: Akkaya, Cem; Conrad, Alexander; Wiebe, Janyce M. & Mihalcea, Rada, 1974-
Partner: UNT College of Engineering

Word Sense and Subjectivity

Description: This paper discusses word sense and subjectivity.
Date: July 2006
Creator: Wiebe, Janyce M. & Mihalcea, Rada, 1974-
Partner: UNT College of Engineering

Characterizing Humour: An Exploration of Features in Humorous Texts

Description: This paper investigates the problem of automatic humor recognition, and provides an in-depth analysis of two of the most frequently observed features of humorous text: human-centeredness and negative polarity. Through experiments performed on two collections of humorous texts, the authors show that these properties of verbal humor are consisted across different data sets.
Date: February 2007
Creator: Mihalcea, Rada, 1974- & Pulman, Stephen
Partner: UNT College of Engineering