Karahalios to receive College of Engineering research award
CS faculty member Karrie Karahalios was one of those selected to receive a Dean’s Award for Excellence in Research for Associate Professors, which recognizes the best University of Illinois College of Engineering research conducted within the last five years by mid-career faculty. Karahalios’ work in social computing explores how computer algorithms shape today’s world, as well as the ethics behind them.
While companies and organizations use algorithms to create customer product recommendations, personalized search results, instant credit approvals, and automated trading systems, society is largely unaware of the algorithms’ power.
In 2014, Karahalios set out to examine how users’ experiences were being shaped on Facebook, asking users about their perception of the Facebook News Feed curation algorithm. She discovered that 62.5 % of participants were not aware of the News Feed curation algorithm’s existence at all—and that it had serious consequences on their real relationships when participants used News Feed to make inferences about them.
She also developed a system, FeedVis, to reveal the differences between the curated feed and an unadulterated feed, and she studied how users perceive the difference. After initially feeling betrayed, over time, many users’ knowledge about the algorithm actually increased their satisfaction with the product.
Karahalios’ landmark study, “I always assumed that I wasn't really that close to [her]: Reasoning about Invisible Algorithms in News Feeds,” was published in 2015 and won a Best Paper Award at CHI, the top conference in the field of human-computer interaction. This work also attracted a tsunami of media attention—with articles in Time, The HuffingtonPost, Fortune, The Christian Science Monitor, and The International Business Times—helping to spark an important discussion about algorithmic manipulation.
In a 2014 paper, Karahalios and her co-authors introduced the concept of algorithm audits to detect how Internet platforms are treating their users. In May 2016, the White House’s report on Big Data and Civil Rights named “algorithm auditing” as one of five national priorities.
Subsequently, Karahalios has published a number of studies examining algorithms: arguing for an “algorithmic ethics,” understanding the effects of algorithm awareness, quantifying search bias for political searches in social media, and proposing a research method for detecting discrimination on Internet platforms.
Because most algorithms are in private hands, with provisions of the Computer Fraud and Abuse Act (CFAA) being a key challenge to study, Karahalios has become one of several researchers collaborating with the American Civil Liberties Union in a lawsuit challenging the act, in order to allow for the collection of data to enable algorithm audits.
During the past five years, Karahalios has published 37 peer-reviewed papers, 2 book chapters, and one book, frequently in the very top venues of her field (CHI, CSCW, and UIST). She’s also written an article about the Facebook News Feed for MIT Technology Review and co-authored an op-ed (about the CFAA) in The Guardian.
She was awarded a 2015 NCSA Faculty Fellowship and has received research grants from Facebook and Adobe. In total, she has brought in more than $1.8 million in research funding, primarily from the NSF.
Karahalios will receive the Dean’s Award in late April at the Engineering at Illinois faculty awards ceremony.