Social Deliberative Skills

PEER REVIEWED Publications

Xu, X., Murray, T., Woolf, B.P. & Smith, D. ( 2014). Identifying Social Deliberative Behavior from Online Communication -- A Cross-domain Study. Proceedings of FLAIRS: Florida Artificial Intelligence Research Society Conference, May, 2014. Pensacola Beach, Florida.
Xu, X., Murray, T., Woolf, B. P., & Smith, D. A. (2014). Social network signatures of effective online communication. In Trausan-Matu, S., Boyer, K., Crosby, M., & Panourgia, K. (Eds.). (2014). Intelligent Tutoring Systems: 12th International Conference, ITS 2014, Honolulu, HI, USA, June 5-9, 2014. Proceedings (Vol. 8474). Springer.
Murray, T., Wing, L., Woolf, B., Wise, A., Wu, S., Clarke, L. Osterweil, L., Xu, X. (2013). A Prototype Facilitators Dashboard: Assessing and visualizing dialogue quality in online deliberations for education and work. Proceedings of The 2013 International Conference on e-Learning, e-Business, Enterprise Information Systems, and e-Government (EEE-2013). Las Vegas, July 2013. (slides) (pdf) (abstract)
Xu, X., Murray, T. , & Woolf, B. (2013). Text Analysis of Deliberative Skills in Undergraduate Online Dialogue: Using L1 Regularized Logistic Regression to Model Psycholinguistic Features. Proceedings of The 2013 International Conference on e-Learning, e-Business, Enterprise Information Systems, and e-Government (EEE-2013). Las Vegas, July 2013. (slides) (pdf) (abstract)
Murray, T., Stephens, A.L., Woolf, B.P., Wing, L., Xu, X., & Shrikant, N. (2013). Supporting Social Deliberative Skills Online: the Effects of Reflective Scaffolding Tools. Proceedings of 5th International Conference on
Online Communities and Social Computing (eSociety at HCII 2013). Las Vegas, July 2013. (slides) (pdf) (abstract)
Murray, T., Xu, X. & Woolf, P.B. (2013). An Exploration of Text Analysis Methods to Identify Social Deliberative Skills. In Proceedings of 16th International Conference on Artificial Intelligence in Education (AIED-2013). Memphis, TN, July 2013. K. Yacef et al. (Eds.): LNAI 7926, pp. 811–814. (short submission pdf; Extended version pdf) (abstract)
Xu, X., Murray, T., Smith, D. & Woolf, B.P. (2013). Mining Social Deliberation in Online Communication: If You Were Me and I Were You. Proceedings of Educational Data Mining (EDM-2013). Memphis, TN, July, 2013. (pdf) (abstract)
Murray, T., Wing, L., Woolf, B., (2013). A Dashboard for Visualizing Deliberative Dialogue in Online Learning. Proceedings of 2nd Workshop on Intelligent Support for Learning in Groups—in association with AIED 2013 (Kim & Kumar Eds.). July, 2013, Memphis, TN, USA. (pdf) (abstract)
Murray, T. (2013). Toward Defining, Justifying, Measuring, and Supporting Social Deliberative Skills. Proceedings of Workshop on Self Regulated Learning — in association with AIED 2013 (Weerasinghe, du Boulay, & Biswas Eds.). July, 2013, Memphis, TN, USA. (pdf) (abstract)
Shrikant, N. (2013). The trajectory of resistance to authority in online academic institutional talk. Proceedings of International Communication Association. London, June 17-21, 2013. (pdf) (abstract)
Murray, T., Woolf, B., Xu, X., Shipe, S., Howard, S. & Wing, L. (2012). Supporting social deliberative skills in online classroom dialogues: Preliminary results using automated text analysis. Proceedings of 11th International Conference on Intelligent Tutoring Systems (ITS-2012). S.A. Cerri and B. Clancey (Eds.). LNCS 7315, pp. 669–671, 2012. June 2012, Chania, Greece. (short subm. pdf) (extended paper pdf) (abstract)
Murray, T., Woolf, B., Xu, X., Shipe, S., Howard, S. & Wing, L. (2012). Towards Supporting Social Deliberative Skills in Online Group Dialogues. Presented at The 7th Annual Interdisciplinary Network for Group Research Conference (InGroup). Chicago July 12-14, 2012.
Xu, X., Smith, D., Murray, T. & Woolf, B. (2012). Analyzing Conflict Narratives to Predict Settlements in EBay Feedback Dispute Resolution. Proceedings of the 2012 International Conference on Data Mining (DMIN12), July, 2012, Las Vegas. (pdf) (abstract)
Woolf, B.P., Murray, T., Xu, X., Osterweil, L., Clarke, L., Wing, L., Katsh, E. (2012). Annotations and Computational Predictors in Online Social Deliberation. Presented at the 6th International AAAI Conference On Weblogs And Social Media (ICWSM). June 4, Dublin, Ireland. (pdf) (abstract)
Murray, T. (2009). Online Curriculum and Dialog Design for Ethics Skills for Science and Engineering Students. In Proceedings of World Conference on E-Learning in Corporate, Government, Healthcare, and Higher Education (Vol. 2009, No. 1, pp. 555-564). (pdf)
Murray, T. (2007). Toward collaborative technologies supporting cognitive skills for mutual regard. In Proceedings of the 8th International Conference on Computer Supported Collaborative Learning (pp. 542-544). International Society of the Learning Sciences. (extended version pdf)
Murray, T. (2003). Toward supporting information quality in rhetorical, dialogic, and collective on-line communication. In Proceedings of Workshop on Metacognition and Self-regulation in Learning with Metacognitive Tools. R. Azevedo (Ed.).

ABSTRACTS:

Murray, T., Xu, X. & Woolf, P.B. (2013). An Exploration of Text Analysis Methods to Identify Social Deliberative Skills. In Proceedings of 16th International Conference on Artificial Intelligence in Education (AIED-2013). Memphis, TN, July 2013. K. Yacef et al. (Eds.): Springer LNAI 7926, pp. 811–814.

We report on text processing and machine learning methods with the goal of building classifiers for social deliberative skill, i.e. the capacity to deal productively with heterogeneous goals, values, or perspectives. Our corpus includes online deliberative dialogue from three diverse domain contexts. We use the LIWC and CohMetrix linguistic analysis tools to generate feature sets for machine learning. We report on our evaluation of various machine learning algorithms, feature selection methods, and cross-domain training methods.

Xu, X., Murray, T., Smith, D. & Woolf, B.P. (2013) . Mining Social Deliberation in Online Communication: If You Were Me and I Were You. Proceedings of Educational Data Mining (EDM-2013). Memphis, TN, July, 2013.

Social deliberative skills are collaborative life-skills. These skills are crucial for communicating in any collaborative processes where participants have heterogeneous opinions and perspectives driven by different assumptions, beliefs, and goals. In this paper, we describe models using lexical, discourse, and gender demographic features to identify whether or not participants demonstrate social deliberative skills from various online dialogues. We evaluate our models using three different corpora with participants of different educational and motivational levels. We propose a protocol about how to use these features to build models that achieve the best in-domain performance and identify the most useful features for building robust models in cross-domain applications. We also reveal lexical and discourse characteristics of social deliberative skills.
Murray, T., Stephens, A.L., Woolf, B.P., Wing, L., Xu, X., & Shrikant, N. (2013). Supporting Social Deliberative Skills Online: the Effects of Reflective Scaffolding Tools. Proceedings of 15th International Conference on Human-Computer Interaction (HCII-2013). Las Vegas, July 2013.

We investigate supporting higher quality deliberations in online contexts by supporting what we call "social deliberative skills," including perspective-taking, meta-dialog, and reflecting on one's biases. We report on an experiment with college students engaged in online dialogues about controversial topics, using discussion forum software with "reflective tools" designed to support social deliberative skills. We find that these have a significant effect as measured by rubrics designed to asses dialogue quality and social deliberative behaviors.
Murray, T., Wing, L., Woolf, B., Wise, A., Wu, S., Clarke, L. Osterweil, L., Xu, X. (2013). A Prototype Facilitators Dashboard: Assessing and visualizing dialogue quality in online deliberations for education and work. Proceedings of The 2013 International Conference on e-Learning, e-Business, Enterprise Information Systems, and e-Government (EEE-2013). Las Vegas, July 2013.

The emerging next generation ("Web 3.0") of socio-technological tool development is adding additional support for reflecting on and improving the quality of online information, communication, and action coordination. An important opportunity is that online systems can include tools that directly support participants in having higher quality and more skillful engagements. We are evaluating dialogue software features that support participants directly and "dashboard" tools that support third parties (mediators, teachers, facilitators, moderators, etc.) in supporting higher quality deliberation. In this paper we will focus on our work in educational settings (college classes) and on our development of a Facilitators Dashboard that visualizes dialogue quality indicators for use as facilitation tools or participant social awareness tools. We are particularly interested in supporting the "social deliberative skills" that interlocutors need to build mutual understanding and mutual regard in complex or contentious situations.
Xu, X., Murray, T. , & Woolf, B. (2013). Text Analysis of Deliberative Skills in Undergraduate Online Dialogue: Using L1 Regularized Logistic Regression to Model Psycholinguistic Features. Proceedings of The 2013 International Conference on e-Learning, e-Business, Enterprise Information Systems, and e-Government (EEE-2013). Las Vegas, July 2013.

We report on a text analysis and machine learning study of social deliberative skill using online dialogues on controversial topics from a college class. We report on our comparison between using the LIWC and Cohmetrix text analysis feature sets, as well as demographic feature information in an L1 Regularized Logistic Regression machine learning algorithm.
Murray, T. (2013). Toward Defining, Justifying, Measuring, and Supporting Social Deliberative Skills. Proceedings of Workshop on Self Regulated Learning — in association with AIED 2013 (Weerasinghe, du Boulay, & Biswas Eds.). July, 2013, Memphis, TN, USA

Social deliberative skill is the capacity to deal productively with heterogeneous goals, values, or perspectives, especially those that differ from ones own, in deliberative situations. In other papers we describe our team's initial results in exploring this domain, which includes evaluating software features hypothesized to support SD-skills in participants, using machine learning and text analysis methods to recognize SD-skills and other indicators of deliberative quality, and prototyping a Facilitators Dashboard to help third parties get a birds-eye-view of important aspects of an online deliberation so that they can better help participants bring SD-skills to bear within dialogues on controversial topics. In this paper we take the opportunity to expand upon the nature and importance of SD-skills as we currently understand them at a more theoretical level.
Murray, T., Wing, L., Woolf, B., (2013). A Dashboard for Visualizing Deliberative Dialogue in Online Learning. Proceedings of 2nd Workshop on Intelligent Support for Learning in Groups—in association with AIED 2013 (Kim & Kumar Eds.). July, 2013, Memphis, TN, USA.

New and emerging online trends in group education, work and communication have lead to a dramatic increases in the quantity of information and connectivity without always supporting—and sometimes sacrificing—their quality. An important opportunity is that online systems can include tools that directly support participants in having higher quality and more skillful engage- ments. We are evaluating dialogue software features that support participants directly and "dashboard" tools that support third parties (mediators, teachers, facilitators, moderators, etc.) in supporting higher quality deliberation. We will focus on our work in educational settings (college classes) and on our develop- ment of a Facilitators Dashboard that visualizes dialogue quality indicators for use as facilitation tools or participant social awareness tools. The Dashboard makes use of text analysis methods to highlight indicators of dialogue quality. We are particularly interested in supporting the "social deliberative skills" that interlocutors need to build mutual understanding and mutual regard in complex or contentious situations.
Shrikant, N. (2013). The trajectory of resistance to authority in online academic institutional talk. Proceedings of International Communication Association. London, June 17-21, 2013.

The ambiguous hierarchy existing in academia is a source of tension in academic discussions, where deliberation is encouraged, but those who rank highly are more likely to control the decision making process. This paper takes a conversation analysis (CA) approach to analyze online academic interactions among an Advisory Committee formed, in part, to solve a conference scheduling issue. This analysis will examine how participants invoke and negate hierarchy during these interactions. Robert G, the appointed leader of the listserv discussion group, OrgE, consistently tries to control the conversation and make decisions without the input of OrgE members. OrgE members resist Robert's control by constructing strong disagreements, negative assessments, and performing intersubjectivity work. This paper presents Robert's first email to the group and then follows by presenting three of the many resistance episodes to Robert. These emails illustrate the increasing hostility participants express, which leads to their resignation, causing a breakdown in communication on the listserv.
Xu, X., Smith, D., Murray, T. & Woolf, B. (2012). "Analyzing Conflict Narratives to Predict Settlements in EBay Feedback Dispute Resolution." Proceedings of the 2012 International Conference on Data Mining (DMIN12), July, 2012, Las Vegas.

We explore the possibility of predicting settlements in online disputes by performing text-analysis on conflict narratives from disputant parties. The experiment domain is eBay Motor vehicles, in which disputants try to resolve complaints, possibly working with online human mediators. The conflict discourse is analyzed based on the divergence of topic distributions in a generative model extending Latent Dirichlet Allocation (LDA) to include role information. A set of distance schemes and a heuristic are designed for various negotiation scenarios to predict settlements. We analyze the quality of discovered topics in terms of topic coherence and evaluate settlement classification and prediction power for settlements on unseen data. Experimental results show that this unsupervised model outperforms a state-of-the-art supervised learner on precision, recall, and F-measure. The performance of a supervised learner with derived features from this model outperforms that using bag-of-features in terms of precision and efficiency.
Murray, T., Woolf, B., Xu, X., Shipe, S., Howard, S. & Wing, L. (2012). "Supporting social deliberative skills in online classroom dialogues: Preliminary results using automated text analysis." Proceedings of 11th International Conference on Intelligent Tutoring Systems (ITS-2012). S.A. Cerri and B. Clancey (Eds.). LNCS 7315, pp. 669–671, 2012. June 2012, Chania, Greece.

We present results from a study in which we tested features of online dialogue software meant to scaffold "social deliberative skills," which include social perspective-taking, question-asking, meta-dialog, and reflecting on how one's biases and emotions are impacting a dialogue. Social deliberative skills are important capacities in a wide array of social contexts in which people with differing goals, values, or perspectives deliberate toward some end, including civic engagement and dispute resolution. In this study we look at online dialogue on controversial topics in a college classroom. In addition to hand coding of the dialogue text we are exploring the use of automated text analysis tools (LIWC and Coh-Metrix) to identify relevant features. Automated analysis might allow for adaptive or intelligent scaffolding of dialogue software features, and could also be used in a Facilitator Dashboard, which we are now prototyping, to bring a facilitator's attention to critical junctures in deliberative dialogues. In our preliminary analysis we found suggestive evidence that LIWC-based automated text analysis can differentiate the use of reflective tools and also differentiate some aspects of higher quality deliberative dialogue. In addition to the empirical results, this study contributes to a theoretical framework for the study and support of social deliberative skills.
Woolf, B.P., Murray, T., Xu, X., Osterweil, L., Clarke, L., Wing, L., Katsh, E. (2012). "Annotations and Computational Predictors in Online Social Deliberation." Presented at the 6th International AAAI Conference On Weblogs And Social Media (ICWSM). June 4, Dublin, Ireland.

This research seeks to identify online participants' disposi- tion and skills. A prototype dashboard and annotation scheme were developed to support facilitators and several computational predictors were identified that show statisti- cally significant correlations with dialogue skills as ob- served by human annotators.

PEER REVIEWED Publications

Other Papers and Presentations:

ABSTRACTS: