About Me
After finishing my Ph.D, I started a post-doctoral position at the Learning Research and Development Center, at the University of Pittsburgh. The focus of my post-doctoral research was research and development for the SWoRD Peer-Review writing support system. We developed a revision planning intervention which allowed students to collect the feedback they were receiving and make concrete decisions about how to incorporate the feedback into their drafts. As with my graduate work, I developed softwware, participated in in-class observational studies, extracted and curated data, and performed the final inferential data analysis and evaluation. I received a second post-doctoral research associate positiont with the Center for Instructional Design and Distance Education. This second project was focused on behavioral analysis of student interactions within PITT Online and Coursera courses.
Since leaving academia and entering industry, I've worked as a Data Scientist at Pearson, and then ETS. I've been involved in the end-to-end development of capabilities and prototypes of novel AI-enabled learning experiences, focusing on writing and language learning. I've lead multi-disciplinary teams of software engineer, data scientists, data engineers and data scientists, to build prototypes, develop capabilities, and design measurements which can lead to actionable insights.
Research Projects
An essay scoring prototype which collaborates with an instructor to build an essay scoring model over the course of a semester using online active learning. Contributions include an active learning algorithm for suggesting the order in which the instructor should grade the essays and a UX framework for communicating online learning flow and status to a non-ML fluent user.
UIBK 2019 LAS 2019
P e n d i n g
An analysis of student interactions, including video interactions, in-class quiz interactions, offline assignments completed, and forum posts.
ILS 2018 EDM 2016
A revision planning intervention that allows students to consider how to respond to the feedback received and lessons learned from their reviewing tasks.
IS 2017 EDM Workshops 2016 EDM 2015
Modeling student problem solving processes within an intelligent game-based learning environment for middle-grade science education.
AIIDE 2016 UMAP 2014 INT 2013 INT 2012 CMN 2012
A system for selecting dialogue utterances based on features of the problem progress and task for an intelligent tutorial dialogue system for introductory Java concepts.
An intelligent creativity-enhancement environment for 6th grade writing education. Students are guided through the process of planning and writing a fable, which is then visualized in a storyboard to aid in the creative revision process.
Developed a description logic framework for representing hierarchical actions for both recognition and planning.
Publications & Patents
Capstones & Dissertations
-
Ph.D Dissertation Discovery-Based Goal Recognition in Interactive Narrative Environments Department of Computer Science, NC State University
Advisor: Dr. James Lester
-
Senior Thesis Partial Recognition and Reasoning about Actions Department of Computer Science, University of Illinois
Advisor: Dr. Eyal Amir
Journal Articles
-
Participating by activity or by week in MOOCs. Information and Learning Science 119(9/10), 572-585.
, Carrie Demmans Epp, Christian D. Schunn. (2018) -
Fuhui Zhang, Christian D. Schunn, Charting the routes to revision: An interplay of writing goals, peer comments, and self-reflections from peer reviews. Instructional Science 45(5), 679-707.
. (2017)
Conference Publications
-
Scott Hellman, Mark Rosenstein, Andrew Gorman, William Murray, Lee Becker, Scaling Up Writing in the Curricilum: Batch Mode Active Learning for Automated Essay Scoring. Sixth Conference on Learning @ Scale, Chicago, IL. 2019
, Jill Budden, Peter W. Foltz. -
Wookhee Min, A Generalized Multidimensional Evaluation Framework for Player Goal Recognition.. To Appear In Proceedings of the Twelfth Annual AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment, Burlingame, California. 2016
, Bradford Mott, Jonathan Rowe, Barry Liu, Eun Ha, James Lester.. -
Redifining “What” In Analyses of Who Does What in MOOCs. In Proceedings of the Ninth International Conference on Educational Data Mining, Raleigh, NC. pp. 569-570. 2016
, Christian Schunn, Carrie Demmans-Epp, Yanjin Long.. -
Understanding Revision Planning in Peer-Reviewed Writing. In Proceedings of the Eight International Conference on Educational Data Mining, Madrid, Spain. pp. 544 – 548. 2015 [PDF]
, Christian Schunn, Kevin Ashley. -
Generalizability of Goal Recognition Models in Narrative-Centered Learning Environments. In Proceedings of the Twenty First International Conference on User Modeling, Adaptation and Personalization, Aalborg, Denmark. pp. 278 – 289. 2014 [PDF]
, Jonathan Rowe, Bradford Mott, James Lester.. -
Eun Young Ha, Exploring the Effectiveness of Lexical Ontologies for Modeling Temporal Relations with Markov Logic. In Proceedings of the Eleventh Annual SIGDIAL Meeting on Discourse and Dialogue, Tokyo, Japan. pp. 75–79. 2010 [PDF]
, Carlyle J. Licata, Bradford W. Mott, James C. Lester..
Conference Presentations
-
Loretto, Adam, Godley, Amanda, A think-aloud study of adolescent writers’ peer review and revision strategies. Paper presented at the 2015 Literacy Research Association Annual Conference, Carlsbad, CA.. 2015, December
.
Workshop Publications
-
An apprenticeship model for human and AI collaborative essay grading. The First Workshop on User Interactions for Building Knowledge (UIBK), Los Angeles, CA. 2019
, Lee Becker, Jill Budden, Peter Foltz, Andrew Gorman, Scott Hellman, William Murray, Mark Rosentstein. -
Impact of Revision Planning in Peer-Reviewed Writing. To Appear In Proceedings of the Second Workshop on Computer-Supported Peer Review in Education, Raleigh, NC. 2016
, Christian Schunn, Kevin Ashley. -
Improving Goal Recognition in Interactive Narratives with Models of Narrative Discovery Events. In Proceedings of the Sixth Workshop on Intelligent Narrative Technologies, Boston, MA. pp. 2-8. 2013 [PDF]
, Jonathan P. Rowe, Bradford W. Mott, James C. Lester. -
Toward Narrative Schema-Based Goal Recognition Models for Interactive Narrative Environments. In Proceedings of the Fifth Workshop on Intelligent Narrative Technologies, Stanford, CA. pp. 2-7. 2012 [PDF]
, Jonathan P. Rowe, Bradford W. Mott, James C. Lester. -
Towards finding the fundamental unit of narrative: A Proposal for the Narreme. In Proceedings of the 2012 Workshop on Computational Models of Narrative, Istanbul, Turkey. pp. 44-46. 2012 [PDF]
, Rogelio E. Cardona-Rivera. -
Towards a Computational Model of Narrative Visualization. In Proceedings of the Fourth Workshop on Intelligent Narrative Technologies, Stanford, California. pp. 2-9. 2011 [PDF]
, Julius Goth, Christopher M. Mitchel, Eun Young Ha, Bradford W. Mott, James C. Lester. -
Eun Young Ha, NCSU: Modeling Temporal Relations with Markov Logic and Lexical Ontology. In Proceedings of the Fifth International Workshop on Semantic Evaluation, Uppsala, Sweden. pp. 341-344. 2010 [PDF]
, Carlyle Licata, James C. Lester. -
Julius Goth, Exploring Individual Differences in Student Writing with a Narrative Composition Support Environment. In Proceedings of the First NAACL HLT Workshop on Computational Linguistics & Writing (CL&W), Los Angeles, California. pp. 56-64. 2010 [PDF]
, Eun Young Ha, Jonathan Rowe, Bradford Mott, James Lester.
Issued Patents
-
US Patent 11741849, Systems and methods for interface-based machine learning model output customization, August 29 2023
Inventors: Hellman; Scott (Dallas, TX), Murray; William (Fort Collins, CO) Habermehl; Kyle (Niwat, CO),
(Boulder, CO), Budden; Jill (Boulder, CO), Gorman; Andrew (Bloomington, MN), Rosenstein; Mark (Boulder, CO), Becker; Lee (Boulder, CO), Hopkins; Stephen (Huntington Station, NY), Foltz; Peter (Boulder, CO) -
US Patent 11475245, Systems and methods for automated evaluation model customization, October 18 2022
Inventors: Foltz; Peter (Boulder, CO), Rosenstein; Mark (Boulder, CO),
(Boulder, CO), Becker; Lee (Boulder, CO), Hopkins; Stephen (Huntington Station, NY), Budden; Jill (Boulder, CO), Oros; Louis M. (Denver, CO), Habermehl; Kyle (Niwat, CO), Hellman; Scott (Dallas, TX), Murray; William (Bloomington, MN), Gorman; Andrew (Bloomington, MN) -
US Patent 11449762, Real time development of auto scoring essay models for custom created prompts, Semptember 20 2022
Inventors: Rosenstein; Mark (Boulder, CO), Habermehl; Kyle (Niwat, CO), Hellman; Scott (Dallas, TX),
(Boulder, CO), Foltz; Peter (Boulder, CO), Becker; Lee (Boulder, CO), Oros; Louis M. (Denver, CO), Budden; Jill (Boulder, CO), Derr; Marcia (Boulder, CO) -
US Patent 11443140, Systems and Methods for automated machine learning model training for a custom authored prompt, September 13 2022
Inventors: Hellman; Scott (Dallas, TX), Becker; Lee (Boulder, CO), Dawson; Samuel (Bloomington, MN),
(Boulder, CO), Murray; William (Fort Collins, CO), Habermehl; Kyle (Niwat, CO), Foltz; Peter (Boulder, CO), Rosenstein; Mark (Boulder, CO) -
US Patent 11416551, System and method for automated sequencing database generation, September 21 2020
Inventors: Murray; William (Lone Tree, CO),
(Boulder, CO) -
US Patent 10860940, System and method for automated sequencing database generation, December 8 2020
Inventors: Murray; William (Lone Tree, CO),
(Boulder, CO) -
US Patent 10783185, System and method for automated hybrid sequencing database generation, September 22 2020
Inventors: Murray; William (Lone Tree, CO),
(Boulder, CO) -
US Patent 10754899, System and method for sequencing database-based content recommendation, August 25 2020
Inventors: Murray; William (Lone Tree, CO),
(Boulder, CO)
Work & Education History
Work History
-
ETS - AI Labs2021-2024Managing Sr. Data Scientist
- Manage a team of 4 data scientists and 5 data engineers tasked with collecting data from ML-enabled product prototypes, manage data sets for ML model development and efficacy research, and develop data-driven insights to aid in high level decision making. Hired and onboarded 3 new employees across both lines. Mentored 2 conversions to an AI Engineering role, and 3 promotions.
- Develop a workflow using git CI/CD, python and Docker integrated into an AWS data platform that team members used to deliver KPI dashboards in PowerBI and Tableau for 10 prototypes accessed by 35,000 customers. Dashboards serviced internal and executive stakeholders.
- Lead design and development of a centralized AWS data platform to gather and process over 4 million events from prototypes and products and provide data exports of derivative data sets for research and calculate KPIs to service the dashboards.
- Design and implement workflows for cross-functional collaboration to leverage data to improve outcomes for other functional roles, including educational psychology researchers, learning designers, UX researchers, and ML model developers.
-
Pearson - AI Products & Solutions2017-2021Principal Data Scientist2019-2021Senior Data Scientist2017-2019
- Manage a team of 2 software engineers, 3 data scientists, and a UX researcher to deliver novel NLP product prototypes and capabilities, evaluate the capabilities deployed in products, and deliver over 50 assessment models to over 1000 early customers for evaluation.
- Design, analyze and publish the results of a multi-condition study to convey the usability, accuracy, and acceptability of ML-based prototypes and capabilities.
- Manage data collection efforts, extracting data from internally developed prototypes, sql and nosql product databases, and external collaborators. Manage requirements and delivery timelines for a data annotation tool, design of annotation schema, and annotation process and evaluation to produce data for ML model design.
- Communicate results and protect IP through 2 publications and 8 issued patents.
- Manage requirements definition, development process, and timelines to ensure delivery of python flask product prototypes, market evaluation, and integrated classical and large-language NLP models for products and services.
- Lead cross-functional design & development to integrate ML capabilities into internal and external products, collaborating with internal marketing, software development, and product teams.
-
University of Pittsburgh - Learning Research and Development Center20014-2017Post-doctoral Research Associate2016-2017Post-doctoral Research Assisstant2014-2016
- Design, implement, and deliver a new software feature using Java/Swing into a edtech product to support the writing process in 3 studies with over 300 early customers.
- Design and analyze 3 multi-condition studies to test the efficacy of the designed software.
- Collaborate with researchers and professors in Education and Cognitive Psychology, Computer Science, and Law to design research studies and collect data to evaluate the impact of software solutions and develop NLP models.
- Design and execute an analysis plan, using multiple regression approaches and factor analysis, to interpret student behavior in 4 university MOOCs containing over 10,000 students.
- Communicate findings through 3 publications and presentations, and 2 journal publications.
- Mentor 4 undergraduate students for research experience, and 1 graduate researcher to complete a novel research study and publication. Lead 6 local instructors through a data-annotation task to produce high-quality data labels for further analysis and ML model development.
-
North Carolina State University - Center for Educational Informatics2007-2013Research Assistant2009-2013
- Design and implement software components using Unity/C#, java and lisp, and NLP & behavior-based graphical machine learning models using research toolkits for novel ML-enabled product prototypes, and deliver to classrooms with over 100 early users each.
- Design, implement, and analyze multi-condition studies with over 100 users each for the purpose of evaluating the efficacy of ML interventions.
- Communicate results through 9 publications and presentations.
- Participate in cross-functional collaboration with education researchers, software developers, and AI researchers to design research studies and collect, organize and process data from studies for research and ML model development.
- Mentor 2 undergraduate students for research experience and delivery of software components for prototypes.
Teaching Assistant2007-2009- Graded exams and assignments and held office hours for 3 graduate classes: Intro to Artificial Intelligence, Machine Learning, and Graph Theory
- Developed curriculum, classroom materials and assignments, Graded exams and assignments, and delivered all lectures for 2 semesters of an Intro to AI course at the undergraduate level.
-
GrammatechSummer 2007Summer Intern
- Developed software to analyze gdb output in C and Lisp.
-
IntelSummer 2005Summer Intern
- Developed utility software in VisualBasic, VisualC++, and VBA for the Automation Group.
Education
-
Ph.D. Computer Science2007-2013
North Carolina State University
-
B.S. Computer Science2003-2007
University of Illinois at Urbana-Champagin