INFORMS: What inspired you to research this particular topic?

LAPPAS: I have been working on customer reviews for almost a decade. It is my favorite domain by far. Review fraud has always been a popular research topic, but it has now become a mainstream concern. As we discuss in the paper, even governments are now joining the fight against fake reviews. The timing was right for us to present this work.

INFORMS: Did any of your results surprise you?

LAPPAS: Any system has its limits. If you stress it enough, it eventually falls apart. Review-based reputation systems are no different. I knew that, if you inject enough fake reviews, you can manipulate the ranking of any business. What I didn’t expect was that you can actually do serious damage with only a small number of injections. All you have to do is be strategic and inject the right positive or negative reviews to the right business. It’s scary, especially given the influence that these rankings have on users. I would imagine it’s even scarier for business owners.

INFORMS: What is the most important take-away you hope readers will learn from your paper?

LAPPAS: Fake reviews are here to stay. Detection algorithms are getting better but fake reviews are also getting more realistic and harder to detect. If you are a business owner, it is crucial to keep track of your review-based visibility and respond to both positive and negative reviews. Build trust by letting the users know that you care. If you are Yelp or TripAdvisor, you have to go beyond analyzing review text. You should monitor the visibility of all businesses in your database. The quality of a business does not change in just a few days or weeks, so neither should its ranking and visibility. If you are a user, don’t just focus on top-ranked items. Rankings can help, but they can also be manipulated. The difference between #1 and #10 is not as great as you think.

INFORMS: Tell us about the process of writing this paper.

LAPPAS: There were several challenges, such as defining review-based visibility and simulating how users process rankings. Once all the technical components were in place, the process of writing the paper went very smoothly. The review team highlighted several points after the first submission, but I appreciated the fact that they also gave us specific directions on how to address them. They really helped us take our work to a different level.

INFORMS: Have you heard about the site “Fakespot”? Would you trust its analysis and identification of fake reviews?

LAPPAS: I have. I think such websites are useful, as they create awareness about fake reviews. However, I can’t trust a system if I don’t know how it works. Fakespot’s description of their detection engine is rather vague. Their results seem reasonable, but nothing that I haven’t seen before. There are dozens of research papers on this problem, so building a detection engine is not that hard. The real challenge in this domain is the lack of ground truth data, which makes it hard to verify that an engine works. Hiding the details of your algorithm makes verification even harder.

INFORMS: Why was it important for you to publish in Information Systems Research?

LAPPAS: If you are an assistant professor in IS, you dream about publishing in ISR, Management Science, and MISQ. Even though these are all elite journals, I think that ISR is the most intuitive fit for someone with a CS background like myself.

INFORMS: How do you yourself keep up-to-date on the latest research in your field?

LAPPAS: I read as many papers as I can and I try to attend as many conferences as possible. Over the years, I have also built connections with relevant people from industry who help me discover challenging problems with real applications.

INFORMS: What about your career might surprise us?

LAPPAS: My background is in computer science; I’m very new to IS. CS people focus on conferences and care more about methods. IS people target journals and care more about findings. I had to learn the ropes and adjust. Not many people make this transition, so the two communities are disconnected. Hopefully this will change, we have a lot to learn from each other.

INFORMS: Tell us a little about what you are working on now.

LAPPAS: Everyone in the IS and CS research communities is talking about Machine Learning algorithms. However, the average user or business person does not seem to trust such complex “black box” methods that deliver accurate but non-interpretable results. My current work evaluates the hypothesis that we can solve many interesting problems with very simple and intuitive algorithms, as long as we have access to very large datasets. The first problem I have studied in this context is sentiment classification. The results are very promising.

INFORMS: Tell us about your experience presenting at CIST 2016 in Nashville? Did you stay to attend the Annual Meeting? If so, what is your favorite part about that conference?

LAPPAS: I actually gave a couple of talks during the Annual Meeting. The INFORMS crowd is always very diverse, so I value the feedback that I get from them. I got to meet a lot of interesting people. My favorite part was the city. Nashville was a pleasant surprise for me this year. One of the most vibrant cities I have ever visited. I will definitely return soon.

INFORMS: When you’re not using your OR superpowers to try to make the world a better place, what are some of the ways you like to spend your time?

LAPPAS: I like to travel to new places; I try to visit two new cities every year. I also love sports; I watch several games every week.

INFORMS: As an INFORMS member, what benefit do you find most useful?

LAPPAS: The discounted rates and the magazine.

INFORMS: What advice would you give to your younger self?

LAPPAS: Travel as much as you can. The planet is full of amazing places.

INFORMS: Tell us something that not many people know about you.

LAPPAS: I’m an ordained minister and licensed to perform weddings in the state of New York.

INFORMS: Which social network do you use most and why?

LAPPAS: I only use Facebook, it helps me stay in touch with my friends from all over the world.