Curriculum Vitae

TIANYU GU

Marketing Department

Eller College of Management

University of Arizona

1130 East Helen Street

Room 320

Tucson, Arizona 85721

  1201 East Drachman Street

Apartment No. 164

Tucson, Arizona 85719

+1 (520) 442-5440

gutianyu@email.arizona.edu

http://www.gutianyu.com

 

 

EDUCATION

2015~2020 (expected) University of Arizona, USA

Ph.D., Management

Major: Marketing, Minor: Economics

2011~2015 Zhejiang University, China

B.S., Information & Computing Science

 

RESEARCH INTERESTS

Substantive Methodological
Digital Marketing Econometrics
Online Reviews Deep Learning and Statistical Learning
Crowdfunding and Other Internet-Enabled Platforms Natural Language Processing

 

PEER-REVIEWED PUBLICATIONS

Wang, Fei, Wei Chen, Ye Zhao, Tianyu Gu, Siyuan Gao, and Hujun Bao. “Adaptively exploring population mobility patterns in flow visualization.” IEEE Transactions on Intelligent Transportation Systems 18, no. 8 (2017): 2250-2259. 
Wang, Fei, Wei Chen, Feiran Wu, Ye Zhao, Han Hong, Tianyu Gu, Long Wang, Ronghua Liang, and Hujun Bao. “A visual reasoning approach for data-driven transport assessment on urban roads.” In 2014 IEEE Conference on Visual Analytics Science and Technology (VAST), pp. 103-112. IEEE, 2014.

 

WORKING PAPERS

Tianyu Gu, Yong Liu, Madhu Viswanathan, “Differentiation in Online Product Reviews: A Machine Learning Based Analysis”, job market paper, under review at Marketing Science
Tianyu Gu, Bikram Ghosh, Yong Liu, “Effects of Text and Image on Reward-Based Crowdfunding Performance”, being finalized for submission to Journal of Marketing Research
Tianyu Gu, Yong Liu, Junming Yin, “Capturing Virtual Business Opportunities from Real-World Events”, manuscript being prepared for submission to Management Science 
Nooshin L. Warren, Matthew Farmer, Tianyu Gu, Caleb Warren, “How to Write Research Papers That Have a Larger Impact”, being finalized for submission to Journal of Consumer Research

 

CONFERENCE PRESENTATIONS

Understanding Review Differentiation and Its Motivations: A Machine Learning Based Text Analysis of Yelp Reviews

–          China Marketing International Conference, Shanghai, 2018

–          INFORMS Marketing Science Conference, Los Angeles, CA, 2017

Capturing Virtual Business Opportunities from Real-World Events: Findings and Insights from Sports Video Games

–          Wharton Customer Analytics Initiative Symposium, San Francisco, CA, 2017

A Visual Reasoning Approach for Data-Driven Transport Assessment on Urban Roads

–          China Visualization Conference, Beijing, 2014

 

PROFESSIONAL CERTIFICATE

Certificate in College Teaching, University of Arizona (expected 2020)

 

TEACHING EXPERIENCE

BNAD 277: Analytical Methods in Business (Lab)                             Summer 2017

Instructor, Teaching Evaluation: 4.3/5

MKTG 459: Innovation and Product Management                                Spring 2019

Sessional Instructor (Bass Model)

MKTG 559: Innovation and Product Strategies                            Spring 2016-2019

Teaching Assistant

MKTG 459: Innovation and Product Management                       Spring 2016-2019

Teaching Assistant

MKTG 579: Marketing of Innovations                                             Fall 2015-2017

Teaching Assistant

 

DOCTORAL COURSEWORK

Marketing Marketing Theory I Dr. Linda Price
  Marketing Theory II Dr. Merrie Brucks
  Marketing Research Method I Dr. Mrinal Ghosh
  Marketing Research Method II Dr. Jennifer Savary
  Psychological Aspects of Consumer Behavior Dr. Merrie Brucks
  Socio-Cultural Aspects of Consumer Behavior Dr. Melanie Wallendorf
  Marketing Decision Models Dr. Bikram Ghosh
  Analytical Models in Marketing Dr. Yong Liu
  Structural Models in Marketing Dr. Madhu Viswanathan
  Marketing Strategy Dr. Mrinal Ghosh
     
Economics Mathematics for Economists Dr. Mark Walker
  Statistics for Economists Dr. Keisuke Hirano
  Microeconomics I Dr. Asaf Plan
  Econometrics I&II Dr. Tiemen Wouterson
  Empirical Industrial Organization Dr. Mo Xiao
  Theoretical Industrial Organization Dr. Stanley Reynolds
  Dynamic Models in Industrial Organization Dr. Gautum Gowrisankaran
  Empirical Environmental Economics Dr. Ashley Langer
     
Sociology Social Network Research Method Dr. Ronald Breiger

 

REFERENCES

Yong Liu

Professor of Marketing and Eller Professor

Eller College of Management

University of Arizona

yoliu@eller.arizona.edu

+1 (520) 621-9320

 

Madhu Viswanathan

Assistant Professor of Marketing

 

Eller College of Management

University of Arizona

madhu@email.arizona.edu

+1 (520) 621-2885

 

Bikram Ghosh

Associate Professor of Marketing

 

Eller College of Management

University of Arizona

bghosh@email.arizona.edu

+1 (520) 626-3061

 

Junming Yin

Assistant Professor of Management Information Systems

Eller College of Management

University of Arizona

junmingy@email.arizona.edu

+1 (520) 626-2961

Gautam Gowrisankaran

Arizona Public Service Professor of Economics

Eller College of Management

University of Arizona

gowrisankaran@eller.arizona.edu

+1(520)621-2529

 

 

ABSTRACT OF RESEARCH PAPERS

Tianyu Gu, Yong Liu, Madhu Viswanathan

Differentiation in Online Product Reviews: A Machine Learning Based Analysis

Job market paper, under review at Marketing Science

Consumers today create a large volume of online reviews for various products and services. This paper examines a key aspect of reviews and reviewer behavior: whether and how the content of a review systematically differs from earlier reviews of the same product or service. Content differentiation becomes particularly important as more reviews are posted because numerical star ratings tend to converge and there is limited room for a new review and its reviewer to stand out. I apply machine learning (SVM) and deep machine learning (CNN, RNN-LSTM) approaches for natural language processing to restaurant reviews on Yelp.com.

The analysis provides strong evidence for review differentiation along both food and non-food dimensions of restaurants and for reviewers with different levels of reputation (elite vs. non-elite). For both elite and non-elite reviewers, the degree of review differentiation is greater when more reviews are posted. A reviewer differentiates the review more when the star rating associated with the review deviates more from previous star ratings. These findings suggest two important but distinct motivations for review differentiation: 1) to enable the review and the reviewer to stand out from the crowd; 2) to provide support for star ratings. Implications for theory, practice and online platforms are provided.

Tianyu Gu, Bikram Ghosh, Yong Liu

Effects of Text and Image on Reward-Based Crowdfunding Performance

Being finalized for submission to Journal of Marketing Research

Using data scraped from kickstarter.com, I analyze how text and image description of projects affect the performance, i.e. eventual pledged funding, of projects. Crowdfunding has been an active area of research in marketing and related business disciplines. However, the literature has not examined the precise effect of text, image and their synergies on project performance on reward-based crowdfunding platforms. This paper is a step in that direction.

To analyze the unstructured text and image data, I employ a deep learning method based on convolutional neural network (CNN) and transfer learning to extract text characteristics, including readability features and linguistic features, and image characteristics, focusing on two color styles: bright and pastel. These characteristics are then applied to the empirical model to estimate their effects on project performance. The results suggest that both text readability and linguistic features have significant impact on project performance. Moreover, I find significant joint effects between color styles and text effects. My study points to opportunities for crowdfunding entrepreneurs to organize their information strategically to improve the performance of their projects.

Tianyu Gu, Yong Liu, Junming Yin

Capturing Virtual Business Opportunities from Real-World Events: Findings and Insights from Sports Video Games

Manuscript being prepared for submission to Management Science

The virtual goods market has increasingly shaped our business world in the past decade as the social network websites and video game industry have emerged rapidly. However, the uniqueness of virtual goods market, compared with the traditional real-world market, has not been widely investigated. In this research, I investigate two major questions related to the interaction between real-world business and virtual goods market, a) do real-world events impact consumers’ purchase behavior in virtual goods market, b) if so, in what ways and to what extent does the impact occur, and c) is the impact moderated by the characteristics of virtual good consumers?

I obtain the purchase data from a top-tier sports video game, which is licensed by a major sports league and using the brand of the sports league. I use a negative binomial model to investigate the real-world effects, and I find that video game players’ in-game purchase behavior is impacted by the result of real-world sports games. Moreover, I observe that the extent of impact is moderated by the experience of being a video game player. Further, my results show that different categories of real-world events have different impacts on the purchase behavior. In addition, I propose Marked Multi-task Multi-dimensional Hawkes Process as the cross-validation of the econometric model.

Nooshin L. Warren, Matthew Farmer, Tianyu Gu, Caleb Warren

How to Write Research Papers That Have a Larger Impact

Being finalized for submission to Journal of Consumer Research

In this paper, I investigate why consumer research papers lack impact. I argue that one reason is because readers who are not already familiar with the research topic struggle to understand the writing. Experts often forget that their readers don’t know as much about the topic as they do, a phenomenon called the curse of knowledge. The curse of knowledge triggers writing practices—abstract language, a reliance on jargon, passive voice, excessive focus on the literature—that make research papers difficult to understand. Papers that are difficult to understand don’t get cited as much.

Specifically, I use text-mining methods to analyze papers published in the Journal of Consumer Research between 2000 and 2010. I find that papers using concrete language, a lot of examples, active voice, and common words earn more citations than papers using abstract language, few examples, passive voice, and uncommon words. I interpret these results as suggesting that the curse of knowledge, which leads to writing that is abstract, jargon-heavy, and that obscures who is doing what, is preventing many research papers from reaching their potential impact.