Farshad Kooti

Evolution of Conversations in the Age of Email Overload

Email is a ubiquitous communications tool in the workplace and plays an important role in social interactions. Previous studies of email were largely based on surveys and limited to relatively small populations of email users within organizations. In this paper, we report results of a large-scale study of more than 2 million users exchanging 16 billion emails over several months. We quantitatively characterize the replying behavior in conversations within pairs of users. In particular, we study the time it takes the user to reply to a received message and the length of the reply sent. We consider a variety of factors that affect the reply time and length, such as the stage of the conversation, user demographics, and use of portable devices. In addition, we study how increasing load affects emailing behavior. We find that as users receive more email messages in a day, they reply to a smaller fraction of them, using shorter replies. However, their responsiveness remains intact, and they may even reply to emails faster. Finally, we predict the time to reply, length of reply, and whether the reply ends a conversation. We demonstrate considerable improvement over the baseline in all three prediction tasks, showing the significant role that the factors that we uncover play, in determining replying behavior. We rank these factors based on their predictive power. Our findings have important implications for understanding human behavior and designing better email management applications for tasks like ranking unread emails.

Evidence of Online Performance Deterioration in User Sessions on Reddit.

This article presents evidence of performance deterioration in online user sessions quantified by studying a massive dataset containing over 55 million comments posted on Reddit in April 2015. After segmenting the sessions (i.e., periods of activity without a prolonged break) depending on their intensity (i.e., how many posts users produced during sessions), we observe a general decrease in the quality of comments produced by users over the course of sessions. We propose mixed-effects models that capture the impact of session intensity on comments, including their length, quality, and the responses they generate from the community. Our findings suggest performance deterioration: Sessions of increasing intensity are associated with the production of shorter, progressively less complex comments, which receive declining quality scores (as rated by other users), and are less and less engaging (i.e., they attract fewer responses). Our contribution evokes a connection between cognitive and attention dynamics and the usage of online social peer production platforms, specifically the effects of deterioration of user performance.

Analyzing Uber's Ride-sharing Economy

Uber is a popular ride-sharing application that matches people who need a ride (or riders) with drivers who are willing to provide it using their personal vehicles. Despite its growing popularity, there exist few studies that examine large-scale Uber data, or in general the factors affecting user participation in the sharing economy. We address this gap through a study of the Uber market that analyzes large-scale data covering 59 million rides which spans a period of 7 months. The data were extracted from email receipts sent by Uber collected on Yahoo servers, allowing us to examine the role of demographics (e.g., age and gender) on participation in the ride-sharing economy. In addition, we evaluate the impact of dynamic pricing (i.e., surge pricing) and income on both rider and driver behavior. We find that the surge pricing does not bias Uber use towards higher income riders. Moreover, we show that more homophilous matches (e.g., riders to drivers of a similar age) can result in higher driver ratings. Finally, we focus on factors that affect retention and use information from earlier rides to accurately predict which riders or drivers will become active Uber users.

Understanding Short-term Changes in Online Activity Sessions

Online activity is characterized by regularities such as diurnal and weekly patterns, reflecting human circadian rhythms and work and leisure schedules. Using data from the online social networking site Facebook, we uncover temporal patterns at a much smaller time scale: within individual sessions. Longer sessions have different characteristics than shorter ones, and this distinction is already visible in the first minute of a persons session activity. This allows us to predict the ultimate length of his or her session and how much content the person will see. The length of the session and other factors are in turn predictive of when the individual will return. Within a session, the amount of time a person spends on different kinds of content depends on both the persons demographic attributes, such as age and the number of Facebook friends, and the length of the time elapsed since the start of the session. We also find that liking and commenting is very non-uniformly distributed between sessions. Predictions of session duration and activity can potentially be leveraged to more efficiently cache content, especially to mobile devices in places with poor communications infrastructure, in order to improve user online experience.

Twitter Session Analytics: Profiling Users’ Short-Term Behavioral Changes

Human behavior shows strong daily, weekly, and monthly patterns. In this work, we demonstrate online behavioral changes that occur on a much smaller time scale: minutes, rather than days or weeks. Specifically, we study how people distribute their effort over different tasks during periods of activity on the Twitter social platform. We demonstrate that later in a session on Twitter, people prefer to perform simpler tasks, such as replying and retweeting others’ posts, rather than composing original messages, and they also tend to post shorter messages. We measure the strength of this effect empirically and statistically using mixed-effects models, and find that the first post of a session is up to 25 % more likely to be a composed message, and 10–20 % less likely to be a reply or retweet. Qualitatively, our results hold for different populations of Twitter users segmented by how active and well-connected they are. Although our work does not resolve the mechanisms responsible for these behavioral changes, our results offer insights for improving user experience and engagement on online social platforms.

The Social Name-Letter Effect on Online Social Networks

The Name-Letter Effect states that people have a preference for brands, places, and even jobs that start with the same letter as their own first name. So Sam might like Snickers and live in Seattle. We use social network data from Twitter and Google+ to replicate this effect in a new environment. We find limited to no support for the Name-Letter Effect on social networks. We do, however, find a very robust Same-Name Effect where, say, Michaels would be more likely to link to other Michaels than Johns. This effect persists when accounting for gender, nationality, race, and age. The fundamentals behind these effects have implications beyond psychology as understanding how a positive self-image is transferred to other entities is important in domains ranging from studying homophily to personalized advertising and to link formation in social networks.

iPhone's Digital Marketplace: Characterizing the Big Spenders

With mobile shopping surging in popularity, people are spending ever more money on digital purchases through their mobile devices and phones. However, few large-scale studies of mobile shopping exist. In this paper we analyze a large data set consisting of more than 776M digital purchases made on Apple mobile devices that include songs, apps, and in-app purchases. We find that 61% of all the spending is on in-app purchases and that the top 1% of users are responsible for 59% of all the spending. These big spenders are more likely to be male and older, and less likely to be from the US. We study how they adopt and abandon individual app, and find that, after an initial phase of increased daily spending, users gradually lose interest: the delay between their purchases increases and the spending decreases with a sharp drop toward the end. Finally, we model the in-app purchasing behavior in multiple steps: 1) we model the time between purchases; 2) we train a classifier to predict whether the user will make a purchase from a new app or continue purchasing from the existing app; and 3) based on the outcome of the previous step, we attempt to predict the exact app, new or existing, from which the next purchase will come. The results yield new insights into spending habits in the mobile digital marketplace.