Georgia Tech SCL Georgia Tech SCL Georgia Tech SCL

TITLE: Information and Overexposure in Word-of-Mouth Campaigns

ABSTRACT:

Access to information is of major concern in a wide range of domains including health, education, and labor markets. Word-of-mouth campaigns are an effective tool for disseminating information that can be leveraged to improve people’s lives. Within health, specifically, by targeting individuals with crucial health information that might be relevant to them, we can potentially create an information cascade that reaches a large portion of the population, and in turn, improve health outcomes for many people.
 
In traditional models for word-of-mouth campaigns, the objective has generally been based on reaching as many people as possible. However, a number of studies both within health and in commercial settings have shown that the indiscriminate spread of a product by word-of-mouth can result in overexposure where the information or product reaches a part of the population that is not receptive to it. This can lead to negative reputational effects or decrease the efficacy of future campaigns. In the first part of the talk, we ask how we should make use of social influence when there is a risk of overexposure. We develop and analyze a theoretical model for this process; we show that it captures a number of the qualitative phenomena associated with overexposure and provide an efficient algorithm to find the optimal campaigning strategy.
 
In the second part of the talk, we focus on the stark inequalities in the availability of health-related data. We focus on the lack of data on health information needs of individuals in Africa and explore the role that search engines can play in uncovering people’s everyday health information needs, concerns, and misconceptions. We analyze Bing searches related to HIV/AIDS in all 54 nations in Africa, and show that these expose a wide range of themes that shed light on how health information needs vary geographically as well as by demographic groups. We discuss the potential for these results to inform targeted education efforts to decrease the burden of the disease.

BIO: Rediet Abebe is a Ph.D. candidate in Computer Science at Cornell. Her research focuses on algorithms, artificial intelligence, and applications to social good. Her work has generously been supported by fellowships and scholarships through Facebook (2017-2019), Google (2016-2017), and the Cornell Graduate School (2015-2016). She is also a 2013-2014 Harvard Cambridge Scholar.