Swapna Somasundaran

OpinionFinder: A System for Subjectivity Analysis

OpinionFinder is a system that performs subjectivity analysis, automatically identifying when opinions, sentiments, speculations, and other private states are present in text. Specifically, OpinionFinder aims to identify subjective sentences and to mark various aspects of the subjectivity in these sentences, including the source (holder) of the subjectivity and words that are included in phrases expressing positive or negative sentiments.

Recognizing Stances in Online Debates

This paper presents an unsupervised opinion analysis method for debate-side classification, i.e., recognizing which stance a person is taking in an online debate. In order to handle the complexities of this genre, we mine the web to learn associations that are indicative of opinion stances in debates. We combine this knowledge with discourse information, and formulate the debate side classification task as an Integer Linear Programming problem. Our results show that our method is substantially better than challenging baseline methods.

Supervised and Unsupervised Methods in Employing Discourse Relations for Improving Opinion Polarity Classification

This work investigates design choices in modeling a discourse scheme for improving opinion polarity classification. For this, two diverse global inference paradigms are used: a supervised collective classification framework and an unsupervised optimization framework. Both approaches perform substantially better than baseline approaches, establishing the efficacy of the methods and the underlying discourse scheme. We also present quantitative and qualitative analyses showing how the improvements are achieved.

Discourse Level Opinion Interpretation

This work proposes opinion frames as a representation of discourse-level associations which arise from related opinion topics. We illustrate how opinion frames help gather more information and also assist disambiguation. Finally we present the results of our experiments to detect these associations.

Discourse Level Opinion Relations: An Annotation Study

This work proposes opinion frames as a representation of discourse-level associations that arise from related opinion targets and which are common in task-oriented meeting dialogs. We define the opinion frames and explain their interpretation. Additionally we present an annotation scheme that realizes the opinion frames and via human annotation studies, we show that these can be reliably identified.

Opinion Graphs for Polarity and Discourse Classification

This work shows how to construct discourse-level opinion graphs to perform a joint interpretation of opinions and discourse relations. Specifically, our opinion graphs enable us to factor in discourse information for polarity classification, and polarity information for discourse-link classification. This inter-dependent framework can be used to augment and improve the performance of local polarity and discourse-link classifiers.

Content Importance Models for Scoring Writing From Sources

Selection of information from external sources is an important skill assessed in educational measurement. We address an integrative summarization task used in an assessment of English proficiency for nonnative speakers applying to higher education institutions in the USA. We evaluate a variety of content importance models that help predict which parts of the source material should be selected by the test-taker in order to succeed on this task.

Manual Annotation of Opinion Categories in Meetings

This paper applies the categories from an opinion annotation scheme developed for monologue text to the genre of multiparty meetings. We describe modifications to the coding guidelines that were required to extend the categories to the new type of data, and present the results of an inter-annotator agreement study. As researchers have found with other types of annotations in speech data, inter-annotator agreement is higher when the annotators both read and listen to the data than when they only read the transcripts. Previous work exploited prosodic clues to perform automatic detection of speaker emotion (Liscombe et al. 2003). Our findings suggest that doing so to recognize opinion categories would be a promising line of work.

Automated Scoring of Picture-based Story Narration

This work investigates linguistically motivated features for automatically scoring a spoken picture-based narration task. Specifically, we build scoring models with features for story development, language use and task relevance of the response. Results show that combinations of these features outperform a baseline system that uses state of the art speechbased features, and that best results are obtained by combining the linguistic and speech features.

Finding your ``Inner-Annotator'': An Experiment in Annotator Independence for Rating Discourse Coherence Quality in Essays

An experimental annotation method is described, showing promise for a subjective labeling task – discourse coherence quality of essays. Annotators developed personal protocols, reducing front-end resources: protocol development and annotator training. Substantial inter-annotator agreement was achieved for a 4-point scale. Correlational analyses revealed how unique linguistic phenomena were considered in annotation. Systems trained with the annotator data demonstrated utility of the data.