2018- Working Papers: Technology and Information Systems

Expand all

Using retweets when shaping our online persona:  A topic modeling approach, 89 pp.
H. Geva, G. Oestreicher-Singer and M. Saar-Tsechansky
(Working Paper no. 1/2018)
Research no.: 05060100

Theories of personal branding are built on the idea that each individual should be aware of the persona she presents to the world. Nowadays, as social interactions are increasingly shifting to the online arena, users of social platforms are presented with many new opportunities and technology-enabled tools by which they can construct their online personas. A powerful type of tool that has emerged in this ecosystem is the ability to reiterate a friend's activity, that is, to redistribute an exact copy of the content that he or she has posted online (e.g., words, videos, or pictures) and to incorporate it into one's own online image. In this work, we examine how users employ reiteration tools when presenting themselves and shaping their online presence. We focus on retweeting behavior on Twitter and study the spectrum of topics that users choose to reiterate. We hypothesize that users' retweeting behavior will show patterns that are theorized to characterize effective personal branding strategies: Specifically, when reiterating content produced by others, a user will maintain a persona that is consistent with the persona portrayed in her self-produced tweets. We analyze data taken from Twitter over a period of 6 months in 2016, with regard to 3,388 non-expert users and 464 expert users and the users whom they followed. We use LDA topic modeling to derive the topics in each user's self-produced tweets and retweets. We find that users' retweets tend to focus on the topics that they address in their self-produced tweets, instead of adding new topics. Further, we find that a user's retweets do remarkably little to alter the distribution of topics she discusses in her self-produced tweets. Finally, we find that this tendency is more prominent among "expert" users, i.e., professional bloggers who are particularly likely to use Twitter as a personal branding tool. A rigorous identification strategy lends support to the proposition that the observed effects are indeed driven by image-related considerations rather than by alternative factors known to influence retweeting behavior, such as exposure bias (a phenomenon associated with the formation of echo chambers), need for uniqueness, and social dynamics on the Twitter platform.



Online exploration when search topic and popularity ranking are decoupled: Insights on echo chambers, 35 pp.
S. Bar-Gill and N. Gandal
(Working Paper no. 7/2018) 
Research no.: 06080100

Personalized search algorithms produce results that are both topically relevant and ranked by their general popularity and individual fit to users' previous searches and choices. New choices from such tailored lists feed back into the algorithms, over time creating content echo chambers, where content exposure is increasingly biased toward users' and their friends' interests and views. We create an online search environment for TED Talks, where topic and popularity are separately controlled, and study the relationship between users’ characteristics and their reliance on own interests vs. crowd-based popularity sorting in content exploration. In topic-based searches, we randomly block/show popularity information to examine its impact on the tendency to explore. We find that high levels of sociability, previous experience with similar content, and a younger age are associated with an increased susceptibility to echo chambers, represented by little to no exploration and popularity sorting prior to content choice. Opinion leaders may alleviate echo chambers in their social circles as they conduct more topic-based exploration and exhibit lower popularity reliance. Showing popularity information increases opinion leaders' popularity sorting, but does not impact non-leaders' exploration. Our findings identify users' echo chamber risk factors, and suggest that reducing the salience of popularity information may contribute to more balanced content exposure facilitated by opinion leaders.


The effect of novelty on impact and success in academia:  Research in progress, 154 pp.
S. Reichman
(Working Paper no. 10/2018) 
Research no.: 01960100

The big data revolution has transformed more and more areas of business, from banner advertising through product recommendations to sales predictions. Surprisingly, one of the areas that is still lagging behind in adopting analytics is academia. While academic researchers are leading the way in generating new methods and algorithms, when it comes to predicting and ranking academic research, current academic decisions like hiring, tenure and prizes are mostly very subjective. Previous literature, from Kuhn (1962), who discussed the importance of novelty to the development of the scientific process, to a recent work by Boudreau et al. (2014), which shows how innovativeness of research (which they measure as “intellectual distance”) affects the likelihood that it is funded, has discussed the important role of novelty in science. This research aims to continue this stream of literature and to build a set of quantitative novelty measures that could support the academic decision making process. These measurements will be embedded in the models to estimate their predictive efficacy when predicting future academic success of papers and scholars.


Engagement, search goals and conversion – the different m-commerce path to conversion: Research in progress10 pp. 
A. Goldstein, O. Raphaeli and S. Reichman 
(Working Paper no. 11/2018) 
Research No.: 01970100

While the use of smartphones is increasing, conversion rates for mobile platforms are still significantly lower than those for traditional e-commerce channels, suggesting that these platforms are characterized by distinct consumption patterns. In this research, using detailed event log-files of an online jewelry retailer, we analyze user engagement and navigation behaviors on both platforms, model search goals and their effect on purchase decisions, and develop a conversion prediction model. Our initial results show that while user engagement is significantly higher in PC sessions compared to mobile sessions, in buying sessions, mobile sessions reflect a higher level of user engagement than PC sessions. These results indicate that m-commerce involves more than ensuring mobile-compatibility of websites, and that mobile consumers follow a distinct path to purchase involving distinct search and browsing behaviors. Therefore, analysis of the different types of consumption behaviors is necessary to understand the factors that lead to conversion on mobile e-commerce platforms.


The predictive power of engagement in mobile consumption, 20 pp. (in Hebrew)
T. Geva, S.Reichman and  I. Somech 
(Working Paper no. 12/2018)
Research no.: 01980100

One of the prominent segments of mobile commerce is the mobile application market, where consumers download applications from an app store. Importantly, prior work showed that user behavior in mobile settings is substantially different than user behavior in PC settings, and therefore needs to be better understood. In this research, we study for the first time the predictive power of consumer engagement in such mobile settings. We perform both in-sample assessment and predictive capacity evaluation of prediction models of app store conversion based on engagement information. Our findings show that in mobile settings, engagement-based models are highly informative for predicting conversion, and are consistent across different prediction methods. We also estimate the correlation of video and video viewing with installation probability. Results indicate that video is not necessarily an engagement-enhancing feature as usually considered. Engagement analytics may enhance our understanding of the app conversion process and provide accurate purchase decision prediction.

Tel Aviv University makes every effort to respect copyright. If you own copyright to the content contained
here and / or the use of such content is in your opinion infringing, Contact us as soon as possible >>