Don’t Let Them Pull Your Leg
January 30, 2019
How can stakeholder manipulations be minimized in internet searches? How can we avoid being exploited by a navigation system in favor of other users? These questions and others are being addressed by Professor Moshe Tennenholtz from the Technion-Israel Institute of Technology’s Davidson Faculty of Industrial Engineering and Management, who presented his work at the Neural Information Processing Systems (NeurIPS, previously named NIPS) Conference, the most prestigious conference for the machine learning community.
In the work he presented in December at NeurIPS, Professor Tennenholtz and his PhD student Omer Ben-Porat, expounded on the technique they developed to minimize the influences of strategic content providers in recommendation systems. Their technique assigns all providers a value that mitigates possible manipulations.
Today, a considerable amount of human activity takes place on the internet, and many of our decisions are made based on searches conducted using search engines. When we search for a product on Google, we begin by reading from the top, i.e., from the first results we obtain, due to the assumption that Google first presents the best possibilities for us. Similarly, we rely on Waze to guide us along the fastest route to our destination.
But these assumptions are not entirely correct.
Firstly, much of the data that reaches Google is fed into the system by parties with vested interests. Website owners want as many people as possible to visit their sites – both to make a purchase and to raise their sites’ rank in Google. This is how, in 2006, the managers of the BMW’s German site brought their site to be the first result of any query for “used car” in German. This index-boosting technique is widely used by many companies, to the tune of more than US$60 billion per year.
Apps such as Waze may present another problem. At times, the app sends drivers, without their knowledge, out as scouts to check out alternative routes. Because the driver is not informed when this happens, the question that arises is whether or not he/she actually received the quickest recommended route, and could ultimately suspect that someone provided a suboptimal route.
Finally, recommendations for a specific product online do not take the existing competition between providers into consideration. To address this issue, PhD student Omer Ben-Porat, under the guidance of Prof. Tennenholtz, developed a new recommendation algorithm that provides a recommendation from company X (on its product) weighted by recommendations of other companies, providing the user with the best recommendation. This is based on game theory tools.
This theory links strategic interests, economics and computations, and manifests by a variety of practical aspects, such as internet advertisement auctions – the main income of companies such as Google and Facebook, which enable free internet.
In the game theory framework, algorithms are developed to receive input from stakeholders – strategic providers, who benefit from the output. Professor Tennenholtz’s group takes this a step further, seeking to isolate the “strategic noise”, i.e., the agent incentives, from the input. Together with his colleague, Prof. Oren Kurland, Prof. Tennenholtz is developing stable search engines, based on both theoretical and empirical research, that filter out possible stakeholder manipulations.
“Certain internet systems,” explained Prof. Tennenholtz, “are highly susceptible to stakeholder manipulations, which has brought us to develop smart algorithms that will replace the classical algorithms and will neutralize these manipulations.”
For more than a century, the Technion-Israel Institute of Technology has pioneered in science and technology education and delivered world-changing impact. Proudly a global university, the Technion has long leveraged boundary-crossing collaborations to advance breakthrough research and technologies. Now with a presence in three countries, the Technion will prepare the next generation of global innovators. Technion people, ideas and inventions make immeasurable contributions to the world, innovating in fields from cancer research and sustainable energy to quantum computing and computer science to do good around the world.
The American Technion Society supports visionary education and world-changing impact through the Technion-Israel Institute of Technology. Based in New York City, we represent thousands of US donors, alumni and stakeholders who invest in the Technion’s growth and innovation to advance critical research and technologies that serve the State of Israel and the global good. Over more than 75 years, our nationwide supporter network has funded new Technion scholarships, research, labs and facilities that have helped deliver world-changing contributions and extend Technion education to campuses in three countries.