51视频

05/13/2020
By Ed Brennen

A dozen undergraduate and graduate students from the Manning School of Business took part in the university鈥檚 23rd annual听Student Research and Community Engagement Symposium, which was held virtually this year because of the coronavirus pandemic.

Of the seven undergraduate business students who participated in the symposium, four did so through the Manning School鈥檚 Business and Entrepreneur Scholar in Training (BEST) program: Katrina Bien-Aime, Katelyn Hasenstab, Emmanuel Moges and Frida Nilsson. Benjamin Geary and Dibren Doe also participated, as did Emma Valentine, who did her project as a UML听Immersive Scholar.

On the graduate side, doctoral students Ali Ahmed, David Greenway, Peiyi Jia, Murtaza Nasir and Roberto Santos all presented their work.

鈥淭his has provided me with new insights into conducting meaningful research that can withstand the 鈥楽o what?鈥� question and that can have broader policy implications,鈥� says Santos, whose project examined the impact of foreign venture capital investment on intellectual property rights. 鈥淚t鈥檚 about research with purpose, and not just research for the sake of research.鈥�

鈥淚t鈥檚 about research with purpose, and not just research for the sake of research.鈥� -Ph.D. student Roberto Santos

In lieu of traditional poster presentations, more than 100 UML students submitted 90-second videos to YouTube featuring slides of their work that they narrated with audio or video. For three days, viewers could vote on their favorite projects by 鈥渓iking鈥� the videos. The top vote-getters from each college advanced to the final round, where faculty judges awarded $250 prizes to individual winners and $500 to winning teams.

鈥淚t was wonderful to see such a wide range of compelling research work being done by our students, who did an incredible job of adapting their presentations during these challenging times,鈥� said Manning School Dean听Sandra Richtermeyer.

Richtermeyer thanked faculty advisors for their support of the students鈥� work, and also thanked听 Asst. Profs. Denise Dunlap, Ann Kronrod, Aggie Yuan and Chi Zhang for volunteering as judges.

Chancellor听Jacquie Moloney听announced the winning projects in a live Zoom ceremony on May 1. From the Manning School, Bien-Aime and Nilsson won in the undergraduate category, while Ahmed and Santos won on the graduate side. Here鈥檚 a closer look at their winning projects:

Analyzing the Aspect-Based Sentiment of Consumer Reviews Over Time: Nilsson, a junior accounting major from Lexington, was recognized for her work with Operations and Information Systems (OIS) Assoc. Prof. Amit Deokar to build a machine-learning model that helps companies automatically detect how consumers feel about products, based on text mined from their customer reviews posted online.

Working with 2,500 randomly selected reviews from a Fortune 500 apparel company, Nilsson used aspect-based sentiment analysis, a text analysis technique that broke the reviews down into six categories: layout (including shape and fit), fabric (quality and comfort), newness, packaging, loyalty and durability. She then manually coded each category on a scale of 1 (highly negative) to 5 (highly positive). The results were then used to train several different machine-learning models for automatic sentiment detection, with the best-performing model selected to predict aspect-based sentiment for the entire dataset of reviews.

鈥淲orking directly with an accomplished professor and contributing to work that's discovering new insights was really great,鈥� Nilsson says. 鈥淚 learned about research at a much higher level, and it鈥檚 an experience I would not have had were it not for the BEST program.鈥�

Subscription Business Models and their Effect on Companies鈥� IPOs:听Bien-Aime, a junior marketing and entrepreneurship major from Cambridge, Mass., also conducted her research through the BEST program. Working with Li Sun, an associate professor of marketing, entrepreneurship and innovation (MEI), Bien-Aime looked at whether companies that adopted a subscription business model (think Netflix or Peloton) saw an effect on their IPO performance.听

鈥淎s a future entrepreneur I thought it would be interesting and insightful to learn how different business models could potentially affect a company,鈥� Bien-Aime says.

Analyzing the prospectus and IPO premiums of 273 public firms, Bien-Aime identified 37 that adopted a subscription business model. After running an ordinary least squares regression model, she found that the IPO premiums for companies that adopted subscription models were 9.81 percent higher than for companies that did not 鈥� suggesting that investors provided higher valuation for subscription business models.

鈥淭he project has sparked my interest in research,鈥� Bien-Aime says.听

Firms' Learning Curve in the Software Vulnerability Resolution Process: Ahmed, a doctoral student in business administration (management information systems concentration), wanted to research information systems security but wasn鈥檛 sure where to start. That鈥檚 when his faculty advisor, OIS Asst. Prof. Brian Lee, introduced him to 鈥渂ug bounty鈥� programs: the growing trend of websites, organizations and software developers using crowdsourcing to reward people for finding security vulnerabilities in their systems.

Ahmed decided to analyze the relationship between experience and performance that firms had when resolving security vulnerabilities on bug bounty platforms. Starting last June, he spent six months gathering and cleaning a dataset from HackerOne, a major bug bounty platform. He analyzed the data and found that the time it took for firms to resolve their vulnerability on the platform has a U-shaped relationship with their experience in resolving the reports, meaning that it initially decreases before increasing.

鈥淲e argue that the firms overgeneralize their limited experience initially, which leads to a negative experience effect on resolving performance. However, as the firms encounter more reported vulnerabilities, the actual learning effect dominates and the experience effect benefits the firms' resolving performance,鈥� says Ahmed, a third-year Ph.D. student from Karachi, Pakistan. 鈥淏y understanding this relationship, organizations can estimate their vulnerability resolution performance and predict the time needed to resolve a future vulnerability.鈥�

Ahmed鈥檚 research is particularly timely.

鈥淒ue to COVID-19, it has become more important for an organization to assess the performance and capabilities of their security teams,鈥� says Ahmed, who notes that hackers frequently try to take advantage of a crisis situation.

Foreign Venture Capital Investment and Cross-Border Knowledge Spillovers: Do foreign venture capital (VC) investments in critical U.S. industries lead to intellectual property leaks that could undermine the country鈥檚 competitive advantage? That鈥檚 the question Santos, a third-year doctoral student in business administration (entrepreneurship concentration), explored in his research project with his faculty advisor, Prof. Yi Yang, chair of the MEI department.

鈥淭his is a particularly timely and relevant topic given recent U.S. policy changes that increase the scrutiny of foreign investments in critical U.S. technologies,鈥� says Santos, whose project was accepted for presentation at the 2020 Academy of Management annual meeting in Vancouver, Canada.

Using a 鈥渇ixed-effects Poisson regression analysis,鈥� Santos analyzed a dataset of foreign VC investments in 27 critical U.S. industries and related patenting activity between 2007 and 2017.

鈥淲e found that at low levels of VC investment, knowledge spillovers to a home country increases. But as the frequency of investments increases, knowledge spillovers begin to decrease,鈥� Santos says. 鈥淭hese findings suggest that the U.S. has nothing to fear from knowledge spillovers and should actually encourage foreign VC investment.鈥澨�

Santos, who began the project as part of an independent study with Yang, says the work helped him develop his empirical research and writing skills.