NEW Department of Education funded project:
iCODE: Adaptive Training of Students' Code Comprehension Processes, Amount: $1,999,595; Period: 2022-2023. iCODE (improve source CODE comprehension) is a novel education technology that targets code comprehension, a critical skill for both learners and professionals. Computer Science (CS) and non-CS, e.g., Data Science, and students from underrepresented groups (females, students of color, first generation status) will greatly benefit from iCODE. Furthermore, education materials such as examples for code comprehension activities and assessment instruments will be developed. More news can be found here: https://www.memphis.edu/efiles/icode.php .
I am currently a PI and co-PI on FOUR major NSF grants totaling $7,185,962.00:
My Current NSF Projects:
The Learner Data Institute: Harnessing The Data Revolution To Make The Learning Ecosystem More Effective, Efficient, and Engaging (Currently Funded Amount: $2,584,309.00; 2019-2023; Phase 2: 15 million for 5 years – to be applied for)
https://www.nsf.gov/awardsearch/showAward?AWD_ID=1934745&HistoricalAwards=false).
The goal is to lay the foundations of a Data Science institute whose mission will be to harness the data revolution to further our understanding of how people learn, how to improve adaptive instructional systems (AISs), and how to make emerging learning ecologies that include online and blended learning with AISs more effective, efficient, engaging, and affordable.
Investigating techniques that couple Markov Logic and Deep Learning with applications to discovering strategies to improve STEM learning (Amount: $413,482.00.00, 2020-2023
https://www.nsf.gov/awardsearch/showAward?AWD_ID=2008812&HistoricalAwards=false).
The goal of this project is to develop novel techniques to integrate different but complementary approaches in artificial intelligence (AI). The project combines the strengths of Deep Neural Networks (DNNs) and Markov Logic Networks (MLNs) to address key shortcomings of those techniques when used by themselves.
·Advancing the Science of Learning Data Science with Adaptive Learning for Future Workforce Development, (Amount: $3,439,035.00; 2020-2025
https://www.nsf.gov/awardsearch/showAward?AWD_ID=1918751&HistoricalAwards=false).
This project will advance understanding of how data science is learned by weaving together statistics, programming, and machine learning and experimental results about student learning. It will use this understanding to create an innovative Artificial Intelligence-enabled data science tutor called DataWhys.
Collaborative Research: CSEdPad: Investigating and Scaffolding Students' Mental Models during Computer Programming Tasks to Improve Learning, Engagement, and Retention (Amount: $749,136.00; a collaborative project with Peter Brusilovsky at The University of Pittsburgh whose portion of the budget is $250k; Period: 2018-2023
https://www.nsf.gov/awardsearch/showAward?AWD_ID=1822816&HistoricalAwards=false).
The goal of the project is to extend my intelligent tutoring system DeepTutor for Computer Science education. Specifically, the project will help learners improve their source code comprehension skills.
I also co-PI on the DoD grant totaling $2,300,000.00:
Generalized Intelligent Framework for Tutors (GIFT) Expert Series, funded – Army Research Lab, co-PI, [$2,300,000.00], 2018-2023.
GIFT is an empirically-based, service-oriented framework of tools, methods and standards to make it easier to author computer-based tutoring systems (CBTS), manage instruction and assess the effect of CBTS, components and methodologies. GIFT is being developed under the Adaptive Tutoring Research Science & Technology project at the Learning in Intelligent Tutoring Environments (LITE) Laboratory, part of the U.S. Army Research Laboratory - Human Research and Engineering Directorate (ARL-HRED).
iHATS: Exploring Student Behavior Across Human and Automated Tutoring Systems: Foundations for a Blended Approach
(Role: Principal Investigator; Agency: DoD; Amount: $263,457.00; Period: 2015-2017)
Using existing tools and capabilities, developed in large part previously, we propose a largescale data mining project aimed at extracting useful information and knowledge from the combined data set. In particular, we propose to gain answers to the following research questions.
RQ1. What tutorial tactics, strategies and metastrategies used by human tutors result in the greatest improvement in student learning, as measured by improvement in AMP performance from that prior to the interaction with human tutors to that after the interaction with human tutors? Are there characteristics (such as a particular unit of the curriculum, prior performance, or affective state) that help determine the most appropriate tutorial interaction?
RQ2. What characteristics of student interaction within AMP (like the use of hints) or student affective state (such as frustration or boredom, as inferred from student interactions with AMP) increase or decrease the likelihood that the student would seek assistance from human tutors?
ARL: Generalized Intelligent Framework for Tutors (GIFT)
(Role: Co-Principal Investigator; Agency: Army Research Lab; Amount: $1,289,545; Period: 2013-2018)
GIFT is an empirically-based, service-oriented framework of tools, methods and standards to make it easier to author computer-based tutoring systems (CBTS), manage instruction and assess the effect of CBTS, components and methodologies. GIFT is being developed under the Adaptive Tutoring Research Science & Technology project at the Learning in Intelligent Tutoring Environments (LITE) Laboratory, part of the U.S. Army Research Laboratory - Human Research and Engineering Directorate (ARL-HRED).
Learning from A Big Human Tutoring Service Database
(Role: Principal Investigator; Agency: DoD, Amount: $557,656.00; Period: 2014-2015)
The goal of this project is to automatically discover successful instructional strategies in the form of sequences of tutorial dialogue moves from raw tutorial conversations between professional human tutors and students. We apply unsupervised machine learning techniques to learn from more than 10 million tutorial session transcripts (obtained from a commercial online tutoring service) successful tutorial strategies. These patterns could then be re-used by state-of-the-art educational systems to improve their training effectiveness.
DRK-12: Developing and Testing the Internship-inator, a Virtual Internship in STEM Authorware System
(Role: Co-Principal Investigator; Agency: NSF; Amount: $999,982.00; Period: 2014-2018)
The goal is to develop the Internship-inator as authorware that will enable STEM content developers to design or modify virtual internships to address different audiences, topics, or purposes without requiring significant prior expertise in computer programming or educational game development. This will facilitate research on authorware design for STEM learning tools more generally as well as research on the effects of virtual internship design and implementation on STEM learning.
DeepTutor: An intelligent tutoring system based on deep language and discourse processing and advanced tutoring strategies
(Role: Principal Investigator; Agency: Institute for Education Sciences; Amount: $1,650,272.00; Period: 2010-2014)
DeepTutor is an advanced intelligent tutoring system that fosters students' deep understanding of complex science topics by using recent advances in science education research, called Learning Progressions, and deep natural language and discourse processing techniques. The emphasis on quality interaction and instruction is what sets DeepTutor apart from previous computer tutors with natural language interaction developed during the last few decades. Quality interaction is possible in DeepTutor through the use of a novel, state-of-the-art natural language-based knowledge representation, called the latent semantic logic form or shortly the semantic logic form (SLF), and advanced dialogue management techniques that embed novel conversational goals such as perfect grounding at every turn. Quality instruction in DeepTutor is driven by recent advances in science education research called learning progressions (LPs). LPs capture the natural sequence of mental models and mental model shifts students go through while mastering a topic. DeepTutor can be used as a means to refine and validate LPs. DeepTutor is the first tutoring system that integrates learning progressions.
Based on these theoretical, conceptual, and technological advances, DeepTutor is expected to provide accurate assessment, better communication, and advanced tutoring and instructional strategies; this will result in higher quality interaction between computer tutor and tutee and therefore increased effectiveness on learning gains beyond the interactivity plateau (see Kurt vanLehn's publications on the interactivity plateau).
To learn more about DeepTutor click here.
MetaTutor: Contextual Research - Empirical Research - Detecting, Tracking, and Modeling Cognitive, Affective, and Meta-cognitive Regulatory Processes to Optimize Learning
(Role: Co-Principal Investigator; Agency: NSF; Amount: $1,496.507.00; Period: 2010-2013)
This 3-year grant focuses on examining the effectiveness of using animated pedagogical agents (APAs) as external regulatory agents designed to foster middle school and college students' understanding of complex and challenging science topics (e.g., the circulatory system). Contemporary cognitive and educational research provides evidence that the potential of computer-based learning environments for facilitating learning may be severely undermined by students' inability to regulate several aspects of the learning. For example, students should regulate key cognitive, metacognitive, motivational, social, and affective processes in order to learn about complex and challenging science topics. This research will be conducted in the context of a mixed-initiative intelligent tutoring system called AutoTutor that simulates the discourse patterns and pedagogical strategies of human tutors. The focus of our grant is on conducting interdisciplinary research examining: (1) the role of embedded animated pedagogical agents in collecting data of the complex interactions between cognitive and metacognitive processes during learning about complex science topics with AutoTutor; (2) the effectiveness of animated pedagogical agents as external regulating agents used to detect, trace, model, and foster students' self-regulatory processes during learning about complex science topics with AutoTutor; and (3) the effectiveness of scaffolding methods delivered by animated pedagogical agents in facilitating middle school and college students' self-regulated learning about complex science topics with AutoTutor.
To read more about MetaTutor click here.