Amazon Mechanical Turk for Subjectivity Word Sense Disambiguation
Date: June 2010
Creator: Akkaya, Cem; Conrad, Alexander; Wiebe, Janyce & Mihalcea, Rada
Description: This paper discusses word sense disambiguation. Abstract: Amazon Mechanical Turk (MTurk) is a marketplace for so-called "human intelligence tasks" (HITs), or tasks that are easy for humans but currently difficult for automated processes. Providers upload tasks to MTurk which workers then complete. Natural language annotation is one such human intelligence task. In this paper, the authors investigate using MTurk to collect annotations for Subjectivity Word Sense Disambiguation (SWSD), a course-grained word sense disambiguation task. The authors investigate whether they can use MTurk to acquire good 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. While our results with respect to spammers are inconclusive, the authors are able to obtain high-quality annotations for the SWSD task. These results suggest a greater role for MTurk with respect to constructing a large scale SWSD system in the future, promising substantial improvement in subjectivity and sentiment analysis.
Contributing Partner: UNT College of Engineering
Permallink:digital.library.unt.edu/ark:/67531/metadc31023/
Anchor Nodes Placement for Effective Passive Localization
Date: 2011
Creator: Akl, Robert G.; Pasupathy, Karthik & Haidar, Mohamad
Description: This paper discusses anchor nodes placement for effective passive localization. Abstract: In many applications, the exact location of the sensor nodes is unknown after deployment. Localization is a process used to locate sensor nodes' positional coordinates, which is vital information. The localization is generally assisted by anchor nodes that are also sensor nodes but with known locations. Anchor nodes generally are expensive and need to be optimally placed for effective localization. Passive localization is one of the localization techniques where the sensor nodes silently listen to the global events like thunder sounds, seismic waves, lighting, etc. According to previous studies, the ideal location to place anchor nodes was on the perimeter of the sensor network. This may not be the case in passive localization, since the function of anchor nodes here is different than the anchor nodes used in other localization systems. The authors do extensive studies on positioning anchor nodes for effective localization. Several simulations are run in dense and sparse networks for proper positioning of anchor nodes. The authors show that, for effective passive localization, the optimal placement of the anchor nodes is at the center of the network in such a way that no three anchor nodes ...
Contributing Partner: UNT College of Engineering
Permallink:digital.library.unt.edu/ark:/67531/metadc77116/
Annotating and Identifying Emotions in Text
Date: 2010
Creator: Strapparava, Carlo & Mihalcea, Rada
Description: This book chapter discusses annotating and identifying emotions in text. Abstract: This paper focuses on the classification of emotions and polarity in news headlines and it is meant as an exploration of the connection between emotions and lexical semantics. The authors first describe the construction of the data set used in evaluation exercise "Affective Text" task at SemEval 2007, annotated for six basic emotions: Anger, Disgust, Fear, Joy, Sadness, and Surprise, and for Positive and Negative polarity. The authors also briefly describe the participating systems and their results. Second, exploiting the same data set, the authors propose and evaluate several knowledge-based and corpus-based methods for the automatic identification of emotions in text.
Contributing Partner: UNT College of Engineering
Permallink:digital.library.unt.edu/ark:/67531/metadc31010/
Classifying genes to the correct Gene Ontology Slim term in Saccharomyces cerevisiae using neighbouring genes with classification learning
Date: May 28, 2010
Creator: Amthauer, Heather A. & Tsatsoulis, C. (Costas), 1962-
Description: This article discusses classifying genes to the correct Gene Ontology Slim term in Saccharomyces cerevisiae using neighbouring genes with classification learning. Abstract: Background: There is increasing evidence that gene location and surrounding genes influence the functionality of genes in the eukaryotic genome. Knowing the Gene Ontology Slim terms associated with a gene gives the authors insight into a gene's functionality by informing the authors how its gene product behaves in a cellular context using three different ontologies: molecular function, biological process, and cellular component. In this study, the authors analyzed if they could classify a gene in Saccharomyces cerevisiae to its correct Gene Ontology Slim term using information about its location in the genome and information from its nearest-neighbouring genes using classification learning. Results: The authors performed experiments to establish that the MultiBoostAB algorithm using the J48 classifier could correctly classify Gene Ontology Slim terms of a gene given information regarding the gene's location and information from its nearest-neighbouring genes for training. Different neighbourhood sizes were examined to determine how many nearest neighbours should be included around each gene to provide better classification rules. The authors' results show that by just incorporating neighbour information from each gene's two-nearest neighbours, the ...
Contributing Partner: UNT College of Engineering
Permallink:digital.library.unt.edu/ark:/67531/metadc122144/
A Novel Space Partitioning Algorithm to Improve Current Practices in Facility Placement
Date: March 2011
Creator: Jimenez, Tamara; Mikler, Armin R. & Tiwari, Chetan
Description: This article discusses a novel space partitioning algorithm to improve current practices in facility placement. In the presence of naturally occurring and man-made public health threats, the feasibility of regional bio-emergency contingency plans plays a crucial role in the mitigation of such emergencies. While the analysis of in-place response scenarios provides a measure of quality for a given plan, it involves human judgement to identify improvements in plans that are otherwise likely to fail. Since resource constraints and government mandates limit the availability of service provided in case of an emergency, computational techniques can determine optimal locations for providing emergency response assuming that the uniform distribution of demand across homogeneous resources will yield and optimal service outcome. This paper presents an algorithm that recursively partitions the geographic space into sub-regions while equally distributing the population across the partitions. For this method, the authors have proven the existence of an upper bound on the deviation from the optimal population size for sub-regions.
Contributing Partner: UNT College of Engineering
Permallink:digital.library.unt.edu/ark:/67531/metadc132975/
Computational Models for Incongruity Detection in Humour
Date: March 2010
Creator: Mihalcea, Rada, 1974-; Strapparava, Carlo, 1962- & Pulman, Stephen
Description: This paper discusses computational models for incongruity resolution. Abstract: Incongruity resolution is one of the most widely accepted theories of humor, suggesting that humor is due to the mixing of two disparate interpretation frames in one statement. In this paper, the authors explore several computational models for incongruity resolution. The authors introduce a new data set, consisting of a series of 'set-ups' (preparations for a punch line), each of them followed by four possible coherent continuations out of which only one has a comic effect. Using this data set, the authors redefine the task as the automatic identification of the humorous punch line among all the plausible endings. The authors explore several measures of semantic relatedness, along with a number of joke-specific features, and try to understand their appropriateness as computational models for incongruity detection.
Contributing Partner: UNT College of Engineering
Permallink:digital.library.unt.edu/ark:/67531/metadc31024/
On B.S.E and B.S.ET for the Engineering Profession
Date: 2010
Creator: Barbieri, Enrique; Attarzadeh, Farrokh; Pascali, Raresh; Shireen, Wajiha & Fitzgibbon, William
Description: This article discusses baccalaureate programs for the engineering profession. An educational model for ABET-accredited baccalaureate programs in Engineering (E) and in Engineering Technology (ET) is proposed whereby all students inclined to pursue an engineering career would first complete two years of a 4-year ET program. By the end of the sophomore year, those students interested and skilled enough to follow a more theoretical or conceive-and-design side of an engineering career would go on to complete a degree in perhaps two to four additional years in a department that offered E degrees. The 4-year option would satisfy the Department of Education definition of a 6-year first professional degree. On the other hand, those students interested and skilled enough to follow a more applied or implement-and-operate side of an engineering career would opt to complete a degree in two additional years in a department that offered ET degrees. The model offers clearly defined options to students interested in an industry-based engineering profession two to four years after graduation where conceive-, design-, implement- and operate-tasks are assigned. If adopted, the model will result in several benefits including: (1) improved program marketing; (2) increased enrollment and retention rates; and (3) improved human and facility ...
Contributing Partner: UNT College of Engineering
Permallink:digital.library.unt.edu/ark:/67531/metadc122181/
SemEval-2010 Task 2: Cross-Lingual Lexical Substitution
Date: July 2010
Creator: Mihalcea, Rada; Sinha, Ravi & McCarthy, Diana
Description: In this paper, the authors describe the SemEval-2010 Cross-Lingual Lexical Substitution task, where given an English target word in context, participating systems had to find an alternative substitute word or phrase in Spanish. The task is based on the English Lexical Substitution task run at SemEval-2007. In this article, the authors provide background and motivation for the task, the authors describe the data annotation process and the scoring system, and present the results of the participating systems.
Contributing Partner: UNT College of Engineering
Permallink:digital.library.unt.edu/ark:/67531/metadc31027/
Multilingual Subjectivity: Are More Languages Better?
Date: August 2010
Creator: Banea, Carmen; Mihalcea, Rada & Wiebe, Janyce
Description: This paper discusses multilingual subjectivity. While subjectivity related research in other languages has increased, most of the work focuses on single languages. This paper explores the integration of features originating from multiple languages into a machine learning approach to subjectivity analysis, and aims to show that this enriched feature set provides for more effective modeling for the source as well as the target languages. The authors show not only that they are able to achieve over 75% macro accuracy in all of the six languages they experiment with, but also that by using features drawn from multiple languages they can construct high-precision meta-classifiers with a precision of over 83%.
Contributing Partner: UNT College of Engineering
Permallink:digital.library.unt.edu/ark:/67531/metadc31025/
Quantifying the Limits and Success of Extractive Summarization Systems Across Domains
Date: June 2010
Creator: Ceylan, Hakan; Mihalcea, Rada; Ozertem, Umut; Lloret, Elena & Palomar, Manuel
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. The authors present a study that explores the summary space of each domain via an exhaustive search strategy, and finds the probability density function (pdf) of the ROUGE score distributions for each domain. The authors then use this pdf to calculate the percentile rank of extractive summarization systems. Their results introduce a new way to judge the success of automatic summarization systems and bring quantified explanations to questions such as why it was so hard for the systems to date to have a statistically significant improvement over the lead baseline in the news domain.
Contributing Partner: UNT College of Engineering
Permallink:digital.library.unt.edu/ark:/67531/metadc31026/