LinkedIn
| Scholar
| Vitae
zhexiong [at] cs.pitt.edu
Quote: The most beautiful thing we can experience is the mysterious. It is the source of all true
art and science. Albert Einstein (1879 to 1955)
Hi, I am a Ph.D. student of Computer Science in the Department of Computer Science at University of Pittsburgh. I work with Prof. Diane Litman in PETAL Lab. I received a master of science degree in Computer Science at Emory University and a bachelor degree in Computer Science and Technology at Sichuan University. Prior to my Ph.D. adventure, I was blessed to work with Prof. Jinho Choi in EmoryNLP Lab.
I have a great passion for leveraging natural language processing to build educational applications (such as automated writing and evaluation systems), the intersection of natural language processing and computer vision (such as visual question & answering and multimodal argument mining), and interpretable machine learning approaches (such as neuro-symbolic reasoning and adaptive meta-learning). Please feel free to drop an email at zhexiong [at] cs.pitt.edu if you are interested in discussing or collaborating with me!
News
- 04/15/2023: A paper was accepted to the 19th BEA Workshop at NAACL 2024 in Mexico City, Mexico.
- 04/02/2024: I passed my Oral Ph.D. Comprehensive Exam with flying colors!
- 01/22/2024: A chapter was accepted to Routledge International Handbook on Automated Essay Evaluation.
- 10/23/2023: A paper was accepted to Microsoft Journal of Applied Research (MSJAR) 2023 in Redmond, WA.
- 10/19/2023: A paper was accepted to the 10th ArgMining Workshop at EMNLP 2023 in Singapore.
- 08/17/2023: I will be co-organizing the 11th ArgMining Workshop at ACL 2024 in Bangkok, Thailand.
- 06/05/2023: I started my internship at Microsoft in Redmond, WA.
- 05/23/2023: A paper was accepted to the 18th BEA Workshop at ACL 2023 in Toronto, Canada.
- 05/15/2023: I will organize ImageArg Shared Task in the 10th ArgMining Workshop at EMNLP 2023 in Singapore.
- 04/06/2023: eRevise project awarded $27K Microsoft Azure credits from CloudBank.
- 04/03/2023: A paper was accepted to Northern European Journal of Language Technology (NEJLT) 2023.
- 02/01/2023: I gave a talk in the Computational Linguistics group at University of Groningen in Netherlands.
- 11/19/2022: A paper was accepted to AAAI 2023 in Washington, D.C.
- 11/10/2022: A paper was accepted to Microsoft Journal of Applied Research (MSJAR) 2022 in Redmond, WA.
- 10/17/2022: ImageArg won the Best Paper Award in the 9th ArgMining Workshop at COLING 2022 in South Korea.
- 06/02/2022: A paper was accepted to MICCAI 2022 in Singapore.
- 05/10/2022: I started my internship at Microsoft in Redmond, WA.
- 06/19/2021: I presented our Poster Spotlight paper in the VQA Workshop at CVPR 2021.
- 05/16/2021: A paper was accepted to IoTAIMA 2021.
- 05/17/2021: I started my internship at Amazon Alexa Knowledge Team in Santa Barbara, CA.
- 05/14/2021: Great to join the 175th Emory Commencement for my graduation in Atlanta, GA.
- 05/31/2020: To-do Bar V2.0 that empowers MacOS menu bar was released.
- 03/27/2020: I passed my MS Thesis Defense! Thank you committee members Prof. Jinho Choi, Prof. Shun Yan Cheung, and Prof. Michelangelo Grigni.
Selected Projects
Predicting the Quality of Revisions in Argumentative Writing
This project studies the relationship between Argumentative Contexts (ACs) and Argumentative Revisions (ARs) in argumentative writing. It leverages Chain-of-Thought prompts to facilitate ChatGPT to generate ACs for identifying successful vs. unsuccessful ARs. It contributes valuable insights that open a promising avenue in revision research.
[video] [slides] [poster] [paper] [code] [webpage]
This project studies the relationship between Argumentative Contexts (ACs) and Argumentative Revisions (ARs) in argumentative writing. It leverages Chain-of-Thought prompts to facilitate ChatGPT to generate ACs for identifying successful vs. unsuccessful ARs. It contributes valuable insights that open a promising avenue in revision research.
[video] [slides] [poster] [paper] [code] [webpage]
Task-Adaptive Meta-Learning Framework for Advancing Spatial Generalizability
This project proposed a model-agnostic meta-learning framework that ensembles regionally heterogeneous data into location-sensitive meta tasks through an easy-to-hard task hierarchy. The framework greatly improves model adaptation to a large number of heterogeneous tasks and advances model generalizability.
[video] [slides] [poster] [paper] [code]
This project proposed a model-agnostic meta-learning framework that ensembles regionally heterogeneous data into location-sensitive meta tasks through an easy-to-hard task hierarchy. The framework greatly improves model adaptation to a large number of heterogeneous tasks and advances model generalizability.
[video] [slides] [poster] [paper] [code]
ImageArg: A Multi-modal Tweet Dataset for Image Persuasiveness Mining
This project presented a multimodal dataset consisting of tweet image persuasiveness annotations. The annotations are based on a persuasion taxonomy developed to explore image functionalities and the means of persuasion. The ImageArg was benchmarked with cutting-edge multimodal learning models.
[video] [slides] [paper] [code] [webpage]
This project presented a multimodal dataset consisting of tweet image persuasiveness annotations. The annotations are based on a persuasion taxonomy developed to explore image functionalities and the means of persuasion. The ImageArg was benchmarked with cutting-edge multimodal learning models.
[video] [slides] [paper] [code] [webpage]
Graph-Symbolic VQA with Rich Visual Estimators and No Question-Answer
This project proposed a graph-symbolic framework that mimics human reasoning through a three-stage graph-based approach: 1) image graph and 2) question graph construction, and 3) answering based on symbols in the graphs. Its novel method requires no question-answer label, which can be applied to new domains without collecting annotated data.
[video] [slides] [paper] [poster]
This project proposed a graph-symbolic framework that mimics human reasoning through a three-stage graph-based approach: 1) image graph and 2) question graph construction, and 3) answering based on symbols in the graphs. Its novel method requires no question-answer label, which can be applied to new domains without collecting annotated data.
[video] [slides] [paper] [poster]
Hierarchical Entity Extraction & Ranking with Unsupervised Graph Convolutions
This project leverages semantic and syntactic information within the documents to perform entity extraction and ranking without supervision. In particular, multiple parsing trees are constructed to extract entity mentions in given documents and entity mentions and entity coreference are further employed to build a relation graph.
[slides] [link]
This project leverages semantic and syntactic information within the documents to perform entity extraction and ranking without supervision. In particular, multiple parsing trees are constructed to extract entity mentions in given documents and entity mentions and entity coreference are further employed to build a relation graph.
[slides] [link]
Personal Food-Ordering Recommendation System based on
Conversations
This project implements a personal food recommendation system, empowering Alexa skills that can capture the user’s preference throughout QA utterances. Our system utilizes CBM, UICF, and CBCF to collaboratively build a novel Alexa Skill that provides personal food recommendations.
[video] [slides] [report]
This project implements a personal food recommendation system, empowering Alexa skills that can capture the user’s preference throughout QA utterances. Our system utilizes CBM, UICF, and CBCF to collaboratively build a novel Alexa Skill that provides personal food recommendations.
[video] [slides] [report]
RaspberryPi for Intelligent Printer
This project implements a tiny self-service retriever based on RaspberryPi. We developed multiple integrated systematical ETL components that can automatically extract, transform and load data into the database. Users can retrieve a poem by just clicking a button on the Raspberry board.
[news]
This project implements a tiny self-service retriever based on RaspberryPi. We developed multiple integrated systematical ETL components that can automatically extract, transform and load data into the database. Users can retrieve a poem by just clicking a button on the Raspberry board.
[news]
Selected Publication
- Tianwen Li, Zhexiong Liu, Lindsay Clare Matsumura, Elaine Wang, Diane Litman, Richard Correnti. Using Large Language Models to Assess Young Students' Writing Revisions, In the proceedings of The 19th Workshop on Innovative Innovative Use of NLP for Building Educational Applications (BEA), held in conjunction with The North American Chapter of the Association for Computational Linguistics (NAACL), June 2024. [link]
- Richard Correnti, Elaine Lin Wang, Lindsay Claire Matsumura, Diane Litman, Zhexiong Liu, Tianwen Li. Supporting Students’ Text-based Evidence Use via Formative Automated Writing and Revision Assessment. The Routledge International Handbook of Automated Essay Evaluation, June 2024. [link]
- Zhexiong Liu, Jing Zhang, Jiaying Lu, Wenjing Ma, Joyce Ho. LogicPrpBank: A Corpus for Logical Implication and Equivalence. In the proceedings of The 5th Artificial Intelligence for Education Workshop, held in conjunction with The 38th AAAI Conference on Artificial Intelligence (AAAI-24), February 2024. [link]
- Zhexiong Liu, Cris Benge, Yash Dagli, Siduo Jiang. STIM: Predicting Memory Uncorrectable Errors with Spatio-Temporal Transformer. In the proceedings of Microsoft Journal of Applied Research (MSJAR), January 2024. [link]
- Zhexiong Liu, Mohamed Elaraby, Yang Zhong, Diane Litman. Overview of ImageArg-2023: The First Shared Task in Multimodal Argument Mining, In the proceedings of The 10th Argument Mining Workshop, held in conjunction with the Conference on Empirical Methods in Natural Language Processing (EMNLP), December 2023. [link]
- Zhexiong Liu, Diane Litman, Elaine Lin Wang, Lindsay C. Matsumura and Richard Correnti. Predicting the Quality of Revisions in Argumentative Writing, In the proceedings of The 18th Workshop on Innovative Innovative Use of NLP for Building Educational Applications (BEA), held in conjunction with The Association for Computational Linguistics (ACL), July 2023. [link]
- Zhexiong Liu as part of Authors. NL-Augmenter: A Framework for Task-Sensitive Natural Language Augmentation, In the proceedings of Northern European Journal of Language Technology (NEJLT), April 2023. [link]
- Zhexiong Liu, Licheng Liu, Yiqun Xie, Zhenong Jin, Xiaowei Jia. Task-Adaptive Meta-Learning Framework for Advancing Spatial Generalizability. In the proceedings of The Thirty-Seventh AAAI Conference on Artificial Intelligence (AAAI-23), May 2023. [link]
- Zhexiong Liu, Cris Benge, Stone Jiang. Ticket-BERT: Labeling Incident Management Tickets with Language Models. In the proceedings of Microsoft Journal of Applied Research (MSJAR), January 2023. [link]
- Zhexiong Liu, Meiqi Guo, Yue Dai and Diane Litman. ImageArg: A Multi-modal Tweet Dataset for Image Persuasiveness Mining. In the proceedings of The 9th Argument Mining Workshop, held in conjunction with the International Conference on Computational Linguistics (COLING), October 2022. (Best Paper Award) [link]
- Ke Yu, Shantanu Ghosh, Zhexiong Liu, Christopher Deible, Kayhan Batmanghelich. Anatomy-Guided Weakly Supervised Abnormality Localization in Chest X-rays. In the proceedings of Medical Image Computing and Computer Assisted Intervention (MICCAI), September 2022. [link]
- Zhexiong Liu and Adriana Kovashka. Graph-Symbolic VQA with Rich Visual Estimators and No Question-Answer Labels. In Visual Question Answering Workshop, held in conjunction with the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2021. (Poster Spotlight) [link]
- Hanzhong Zheng, Zhexiong Liu, Dejia Shi. Image-text Discourse Coherence Relation Discoveries on Multi-image and Multi-text Documents. In the proceedings of International Conference on Internet of Things, Artificial Intelligence and Mechanical Automation (IoTAIMA), June 2021. [link]
Experience
- 2023 to 2023: Data & Applied Scientist Intern, Microsoft
Worked with Yash Dagli, Stone Jiang, and Cris Benge - 2022 to 2025: Research Assistant, Pitt PETAL Lab
Work with Prof. Diane Litman - 2022 to 2022: Data & Applied Scientist Intern, Microsoft
Worked with Stone Jiang, Andrew Fogarty, Jason Tongarm, and Cris Benge - 2021 to 2021: Applied Scientist Intern, Amazon Alexa
Worked with Nic Jedema, Manish Gupta, and Ruben Janssen - 2020 to 2021: Graduate Student, Pitt Computer Vision Lab
Worked with Prof. Adriana Kovashka - 2019 to 2020: Research Assistant, Emory NLP Lab
Worked with Prof. Jinho Choi - 2018 to 2019: Graduate Assistant, Emory Rose Library
Worked with Brenna Edwards -
2016 to 2018: Algorithm Engineer, Calabar Information Technology
Worked on machine learning & data mining projects
Education
- 2020 to 2025: Ph.D. Student of Computer Science, University of Pittsburgh, Pittsburgh, PA
- 2021 to 2023: Visiting Student, Cross Registration, Carnegie Mellon University, Pittsburgh, PA
- 2018 to 2020: Master of Science in Computer Science, Emory University, Atlanta, GA
- 2012 to 2016: Bachelor of Engineering in Computer Science & Technology, Sichuan University, China
Teaching
- Human Language Technologies (CS1671, Fall 2021), University of
Pittsburgh
Worked with Prof. Diane Litman - Introduction to Natural Language Processing (CS2731, Fall
2021), University of
Pittsburgh
Worked with Prof. Malihe Alikhani - Operating Systems (CS1550, Spring 2021), University of
Pittsburgh
Worked with Dr. Sherif Khattab - Introduction to Computing for Scientist (CS0011, Spring 2021),
University of
Pittsburgh
Worked with Dr. Paulo Ferreira and Matt Barbosa - Algorithm Implementation (CS1501, Fall 2020, Spring 2022),
University of
Pittsburgh
Worked with Dr. Nicholas Farnan and Dr. Sherif Khattab - Introduction to Statistic Inference (QTM 100, Spring
2020 & Fall 2019),
Emory University
Worked with Dr. Hojin Kim - Theory of Computing (CS424/CS524, Spring 2020), Emory
University
Worked with Prof. Michelangelo Grigni - Graduate Algorithms (CS526, Fall 2019), Emory University
Worked with Prof. Michelangelo Grigni - Mathematics for Quantitative Sciences (QTM 120, Spring
2019), Emory University
Worked with Prof. Frederic Bien
Student Supervision
- Mason Gobat (CS Undergraduate Student)
Academic Service
- Organizer: The 11th ArgMining Workshop at ACL 2024 in Bangkok, Thailand
- ImageArg Shared Task in the 10th ArgMining Workshop at EMNLP 2023 in Singapore
- Reviewer: AAAI 2024, ACL 2024, EDM 2024, CVPR-W 2024, AAAI 2023, ACL 2023, ACL-W 2023
- EMNLP-W 2023, CVPR-W 2023, CVPR-W 2022
- Student Volunteer: AAAI 2024, AAAI 2023, ACL 2023, COLING 2022
Leadership and Community Engagement
- 2022 to 2023: Committee Member, SCI Academic Council, University of Pittsburgh
- 2022 to 2022: Graduate Mentor, Graduate Global Ties Mentor Program, University of Pittsburgh
- 2020 to 2020: Project Mentor, CMU VentureWell Hackathon, Carnegie Mellon University
- 2020 to 2020: Volunteer and Instructor, Senior Citizens Digital Literacy Program, University of Pittsburgh
- 2019 to 2019: Volunteer and Mentor, Emory DataFest Hackathon, Emory University
- 2018 to 2019: Graduate Assistant and Project Developer, Stuart A. Rose Library, Emory University
Selected Awards
- 2024, 2023: Association for Advancement of Artificial Intelligence (AAAI) Student Scholarship
- 2023: Learning Research and Development Center (LRDC) Graduate Student Council Award
- 2023: Association for Computational Linguistics (ACL) Student Travel Award
- 2023: Graduate and Professional Student Government Travel Award
- 2023, 2022: Department of Computer Science Travel Award
- 2022: Best Paper Award in the 9th Argument Mining Workshop at COLING
- 2022: International Conference on Computational Linguistics (COLING) Student Travel Award
- 2020, 2019, 2018: Emory Faculty (Half-Tuition) Scholarship
- 2016: Honors Undergraduate Thesis Program at Sichuan University
- 2016: Honorable Mention at Mathematical Contest in Modeling (MCM)
- 2015: The 3rd Prize at National College Mathematics Competition (CMC)
- 2015: Science and Technology Innovation Award at Sichuan University
- 2015, 2014, 2013: Sichuan University Merit Scholarship
Miscellaneous
- I was born and raised and spent wonderful years in Sichuan in Southwest China.
- I learned Chinese Freehand Landscape Painting for a decade. Check out my Art Works.
Contact
- Zhexiong Liu, he/him/his
- Email: zhexiong [at] cs.pitt.edu
- Mailing Address:
Department of Computer Science
6117 Sennott Square
University of Pittsburgh
Pittsburgh, PA 15260 USA