Tengteng Ma

Tengteng Ma

Assistant Professor

School of Information Systems & Management

Muma College of Business

University of South Florida

Email: tengtengma[at]usf.edu

Biography

I am an assistant professor in the School of Information Systems and Management, Muma College of Business, University of South Florida. I earned my Ph.D. degree in Management Information Systems from University of Illinois at Chicago. My research interest lies in the intersection of Artificial Intelligence and Business Analytics. I study human behaviors on digital platforms using large-scale heterogeneous datasets and design interpretable artificial intelligence models to address complex business problems; I also conduct econometrics-oriented empirical analyses to investigate individual decision making on digital platforms, which are based on the valuable information extracted from semi-structured and unstructured data using machine learning, natural language processing, and computer vision. My studies have won Best Paper Award at WITS 2020, and Best Paper Nominee at HICSS 2021 and ICIS 2021.

Methodologies: Machine Learning, Deep Learning, Natural Language Processing, Computer Vision, Econometrics, Meta-analysis.

Topics: Consumer Behavior, Online Marketing, Recommender System, Social Network, Explainable Machine Learning, Healthcare.

Projects

Graph Neural Network Model with Attention Mechanism for Customer Engagement Prediction

with Yuheng Hu, Yingda Lu, Siddhartha Bhattacharyya

We design a novel Graph Neural Network based deep learning model called GACE to predict customer engagement of brand posts by exploiting large-scale content consumption information from the perspective of heterogeneous networks. In addition, we provide business insights regarding overall brand engagement performance and the change of customer preferences over time, which has practical value for customer relationship management.

Group Size, Content Moderators, and User Engagement in Online Synchronous Platforms

with Keran Zhao, Yili Hong, Yingda Lu, Yuheng Hu

We empirically examine how group size affects viewers’ real-time engagement on online synchronous platforms and how moderators affect this relationship. The findings indicate a congestion effect of increasing group size viewer engagement in the synchronous communication setting and suggest the beneficial role of content moderators.

Content Creator versus Brand Advertiser? The Effect of Inserting Advertisements in Videos on Influencers

with Yingda Lu, Yuheng Hu, Xi Chen, Yuxin Chen

We investigate the impact of inserting commercial endorsements in videos on influencers' popularity and reputation from an omni-dimensional perspective. Leveraging face recognition techniques, we further investigate how this effect can be moderated if content generators demonstrate stronger endorsement by showing their face.

The Role of Dislike Rating in Digital Media Platform: A Natural Experiment

with Ahreum Kim, Yingda Lu, Ali Tafti

We explore the relationship between passive and active engagement activities when one of the engagement channels is removed. By collecting a large-scale dataset from Youtube, we empirically examine the reinforcement and substitution effects between the two types of engagement behaviors.

Modeling Customer Complain Behavior: A Neural Matrix Factorization Approach

with Yuheng Hu, Yingda Lu

We propose a deep learning approach incorporating prior domain knowledge into a principled probabilistic matrix factorization framework to enhance complain behavior predictions and their interpretability.

Characterizing Recommender Systems with User Fatigue Awareness

with Yunjuan Wang, Theja Tulabandhula

We consider an online recommendation setting where platforms need to consider both the positional effects of items and the abandoning behavior of users when they recommend a sequence of items to users. We propose a new Thompson-sampling based algorithm with expected regret that is polynomial in the number of items in this combinatorial setting and show its superior performance in practice.

with Debaleena Chattopadhyay, Hasti Sharifi

We systematically evaluate evidence from controlled studies of interventions using virtual humans on their effectiveness in health-related outcomes. The design and implementation characteristics of these systems are also examined.

Awards

WITS 2020 Best Paper Awards

ICIS 2021 Best Paper Nomination

HICSS 2021 Best Paper Nomination

University of Illinois at Chicago Doctoral Scholarship

University of Illinois at Chicago Graduate College Doctoral Fellowship

Conferences and Workshops

INFORMS Annual Meeting

  • 2022: Invited Presenter
  • 2021: Computational Social Science, Session Chair
  • 2020: Invited Presenter
  • 2019: Invited Presenter

International Conference on Information Systems (ICIS 2021)

Conference on Information Systems and Technology (CIST 2022)

Workshop on Information Technologies and Systems (WITS 2020, 2021)

Hawaii International Conference on System Sciences (HICSS 2021, 2023)

INFORMS Workshop on Data Science (WDS 2020)

Teaching Experience

Instructor

  • IDS 400 Programming for Data Science in Business (2021 Summer - Present)

    Over 70 enrollments per semester (undergraduate + graduate students)

    Average Course Evaluation: 4.5/5

    I have been an independent instructor of this course for five semesters. The course primarily focuses on Python programming skills for data analytics in business, which aims to help students learn the basics of data analytics, and how to program in Python to solve real-world problems.

Teaching Assistant

  • IDS 576 Deep Learning and Applications (2021 Spring - present)
  • IDS 594 Operationalizing Machine Learning (MLOps) (2020 Fall - present)
  • IDS 561 Analytics for Big Data (2019 Fall - present)
  • IDS 566 Advanced Text Analytics for Business (2019 Fall - 2020 Fall)