Early Warnings of Cyber Threats in Online Discussions.
Sapienza, A.; Bessi, A.; Damodaran, S.; Shakarian, P.; Lerman, K.; and Ferrara, E.
In
2017 IEEE International Conference on Data Mining Workshops, ICDM Workshops 2017, New Orleans, LA, USA, November 18-21, 2017, pages 667–674, 2017.
![Early Warnings of Cyber Threats in Online Discussions [link]](//bibbase.org/img/filetypes/link.svg "Paper")
Paper
doi
link
bibtex
8 downloads
@inproceedings{Sapienza2017,
author = {Anna Sapienza and
Alessandro Bessi and
Saranya Damodaran and
Paulo Shakarian and
Kristina Lerman and
Emilio Ferrara},
title = {Early Warnings of Cyber Threats in Online Discussions},
booktitle = {2017 {IEEE} International Conference on Data Mining Workshops, {ICDM}
Workshops 2017, New Orleans, LA, USA, November 18-21, 2017},
pages = {667--674},
year = {2017},
crossref = {DBLP:conf/icdm/2017w},
url = {https://doi.org/10.1109/ICDMW.2017.94},
doi = {10.1109/ICDMW.2017.94},
timestamp = {Thu, 11 Jan 2018 09:07:04 +0100},
biburl = {http://dblp.org/rec/bib/conf/icdm/SapienzaBDSLF17},
bibsource = {dblp computer science bibliography, http://dblp.org}
}
BD2K ERuDIte: the Educational Resource Discovery Index for Data Science.
Ambite, J. L.; Fierro, L.; Geigl, F.; Gordon, J.; Burns, G. A.; Lerman, K.; and Van Horn, J. D
In
Proceedings of the 26th International Conference on World Wide Web Companion, pages 1203–1211, 2017. International World Wide Web Conferences Steering Committee
link
bibtex
@inproceedings{ambite2017bd2k,
title={BD2K ERuDIte: the Educational Resource Discovery Index for Data Science},
author={Ambite, Jos{\'e} Luis and Fierro, Lily and Geigl, Florian and Gordon, Jonathan and Burns, Gully APC and Lerman, Kristina and Van Horn, John D},
booktitle={Proceedings of the 26th International Conference on World Wide Web Companion},
pages={1203--1211},
year={2017},
organization={International World Wide Web Conferences Steering Committee}
}
Graph Filters and the Z-Laplacian.
Yan, X.; Sadler, B. M.; Drost, R. J.; Yu, P. L.; and Lerman, K.
IEEE Journal of Selected Topics in Signal Processing, 11(6): 774–784. 2017.
link
bibtex
abstract
@ARTICLE{Yan2017,
author = {Xiaoran Yan and Brian M. Sadler and Robert J. Drost and Paul L. Yu and Kristina Lerman},
title = {Graph Filters and the Z-Laplacian},
journal = {IEEE Journal of Selected Topics in Signal Processing},
year = {2017},
volume = {11},
number = {6},
pages = {774--784},
abstract = {In network science, the interplay between dynamical processes and the underlying topologies of complex systems has led to a diverse family of models with different interpretations. In graph signal processing, this is manifested in the form of different graph shifts and their induced algebraic systems. In this paper, we propose the unifying Z-Laplacian framework, whose instances can act as graph shift operators. As a generalization of the traditional graph Laplacian, the Z-Laplacian spans the space of all possible Z -matrices, i.e., real square matrices with nonpositive off-diagonal entries. We show that the Z -Laplacian can model general continuous-time dynamical processes, including information flows and epidemic spreading on a given graph. It is also closely related to general nonnegative graph filters in the discrete time domain. We showcase its flexibility by considering two applications. First, we consider a wireless communications networking problem modeled with a graph, where the framework can be applied to model the effects of the underlying communications protocol and traffic. Second, we examine a structural brain network from the perspective of low- to high-frequency connectivity.},
}
In network science, the interplay between dynamical processes and the underlying topologies of complex systems has led to a diverse family of models with different interpretations. In graph signal processing, this is manifested in the form of different graph shifts and their induced algebraic systems. In this paper, we propose the unifying Z-Laplacian framework, whose instances can act as graph shift operators. As a generalization of the traditional graph Laplacian, the Z-Laplacian spans the space of all possible Z -matrices, i.e., real square matrices with nonpositive off-diagonal entries. We show that the Z -Laplacian can model general continuous-time dynamical processes, including information flows and epidemic spreading on a given graph. It is also closely related to general nonnegative graph filters in the discrete time domain. We showcase its flexibility by considering two applications. First, we consider a wireless communications networking problem modeled with a graph, where the framework can be applied to model the effects of the underlying communications protocol and traffic. Second, we examine a structural brain network from the perspective of low- to high-frequency connectivity.
Multi-layer network composition under a unified dynamical process.
Yan, X.; Teng, S.; and Lerman, K.
In
International Conference on Social Computing, Behavioral-Cultural Modeling and Prediction and Behavior Representation in Modeling and Simulation (SBP), pages 315–321, 2017. Springer
link
bibtex
@inproceedings{yan2017multi,
title={Multi-layer network composition under a unified dynamical process},
author={Yan, Xiaoran and Teng, Shang-Hua and Lerman, Kristina},
booktitle={International Conference on Social Computing, Behavioral-Cultural Modeling and Prediction and Behavior Representation in Modeling and Simulation (SBP)},
pages={315--321},
year={2017},
organization={Springer}
}
Language, demographics, emotions, and the structure of online social networks.
Lerman, K.; Marin, L. G.; Arora, M.; de Lima, L. H. C.; Ferrara, E.; and Garcia, D.
Journal of Computational Social Science. Oct 2017.
Paper
doi
link
bibtex
abstract
2 downloads
@ARTICLE{Lerman2017jcss,
author="Lerman, Kristina and Marin, Luciano G.
and Arora, Megha and de Lima, Lucas H. Costa and Ferrara, Emilio and Garcia, David",
title="Language, demographics, emotions, and the structure of online social networks",
journal="Journal of Computational Social Science",
year="2017",
month="Oct",
day="31",
abstract="Social networks affect individuals' economic opportunities and well-being. However, few of the factors thought to shape networks---culture, language, education, and income---were empirically validated at scale. To fill this gap, we collected a large number of social media posts from a major US metropolitan area. By associating these posts with US Census tracts through their locations, we linked socioeconomic indicators to group-level signals extracted from social media, including emotions, language, and online social ties. Our analysis shows that tracts with higher education levels have weaker social ties, but this effect is attenuated for tracts with high ratio of Hispanic residents. Negative emotions are associated with more frequent online interactions, or stronger social ties, while positive emotions are associated with weaker ties. These results hold for both Spanish and English tweets, evidencing that language does not affect this relationship between emotion and social ties. Our findings highlight the role of cognitive and demographic factors in online interactions and demonstrate the value of traditional social science sources, like US Census data, within social media studies.",
issn="2432-2725",
doi="10.1007/s42001-017-0001-x",
url="https://doi.org/10.1007/s42001-017-0001-x"
}
Social networks affect individuals' economic opportunities and well-being. However, few of the factors thought to shape networks—culture, language, education, and income—were empirically validated at scale. To fill this gap, we collected a large number of social media posts from a major US metropolitan area. By associating these posts with US Census tracts through their locations, we linked socioeconomic indicators to group-level signals extracted from social media, including emotions, language, and online social ties. Our analysis shows that tracts with higher education levels have weaker social ties, but this effect is attenuated for tracts with high ratio of Hispanic residents. Negative emotions are associated with more frequent online interactions, or stronger social ties, while positive emotions are associated with weaker ties. These results hold for both Spanish and English tweets, evidencing that language does not affect this relationship between emotion and social ties. Our findings highlight the role of cognitive and demographic factors in online interactions and demonstrate the value of traditional social science sources, like US Census data, within social media studies.
Deep Context: A Neural Language Model for Large-scale Networked Documents .
Wu, H.; and Lerman, K.
In , editor(s),
Proceedings of International Joint Conference on AI (IJCAI), 2017.
link
bibtex
@INPROCEEDINGS{Wu2017,
author = {Hao Wu and Kristina Lerman},
title = {Deep Context: A Neural Language Model for Large-scale Networked Documents },
booktitle = {Proceedings of International Joint Conference on AI (IJCAI)},
year = {2017},
editor = {},
}
Network vector: distributed representations of networks with global context.
Wu, H.; and Lerman, K.
arXiv preprint arXiv:1709.02448. 2017.
link
bibtex
@article{wu2017network,
title={Network vector: distributed representations of networks with global context},
author={Wu, Hao and Lerman, Kristina},
journal={arXiv preprint arXiv:1709.02448},
year={2017}
}
Neighbor-Neighbor Correlations Explain Measurement Bias in Networks.
Wu, X.; Percus, A. G; and Lerman, K.
Scientific Reports, 7(5576). jul 2017.
doi
link
bibtex
@article{Wu2017neighbor,
title={Neighbor-Neighbor Correlations Explain Measurement Bias in Networks},
author={Wu, Xin-Zeng and Percus, Allon G and Lerman, Kristina},
journal={Scientific Reports},
volume={7},
number = {5576},
month = {jul},
year={2017},
publisher={Nature Publishing Group},
doi = {doi:10.1038/s41598-017-06042-0},
}
Dynamics of Content Quality in Collaborative Knowledge Production.
Ferrara, E.; Alipourfard, N.; Burghardt, K.; Gopal, C.; and Lerman, K.
In , editor(s),
Proceedings of 11th AAAI International Conference on Web and Social Media, 2017. AAAI
link
bibtex
abstract
@INPROCEEDINGS{Ferrara2017dynamics,
author = {Emilio Ferrara and Nazanin Alipourfard and Keith Burghardt and Chiranth Gopal and Kristina Lerman},
title = {Dynamics of Content Quality in Collaborative Knowledge Production},
booktitle = {Proceedings of 11th AAAI International Conference on Web and Social Media},
year = {2017},
editor = {},
publisher = {AAAI},
abstract = {We explore the dynamics of user performance in collaborative knowledge production by studying the quality of answers to questions posted on Stack Exchange. We propose four indicators of answer quality: answer length, the number of code lines and hyperlinks to external web content it contains, and whether it is accepted by the asker as the most helpful answer to the question. Analyzing millions of answers posted over the period from 2008 to 2014, we uncover regular short-term and long-term changes in quality. In the short-term,
quality deteriorates over the course of a single session, with each successive answer becoming shorter, with fewer code lines and links, and less likely to be accepted. In contrast, performance improves over the long-term, with more experienced users producing higher quality answers. These trends are not a consequence of data heterogeneity, but rather have a behavioral origin. Our findings highlight the complex interplay between short-term deterioration in performance, potentially due to mental fatigue or attention depletion, and long-term performance improvement due to learning and skill acquisition, and its impact on the quality of user-generated content.},
keywords = {social-behavior},
}
We explore the dynamics of user performance in collaborative knowledge production by studying the quality of answers to questions posted on Stack Exchange. We propose four indicators of answer quality: answer length, the number of code lines and hyperlinks to external web content it contains, and whether it is accepted by the asker as the most helpful answer to the question. Analyzing millions of answers posted over the period from 2008 to 2014, we uncover regular short-term and long-term changes in quality. In the short-term, quality deteriorates over the course of a single session, with each successive answer becoming shorter, with fewer code lines and links, and less likely to be accepted. In contrast, performance improves over the long-term, with more experienced users producing higher quality answers. These trends are not a consequence of data heterogeneity, but rather have a behavioral origin. Our findings highlight the complex interplay between short-term deterioration in performance, potentially due to mental fatigue or attention depletion, and long-term performance improvement due to learning and skill acquisition, and its impact on the quality of user-generated content.
The myopia of crowds: Cognitive load and collective evaluation of answers on Stack Exchange.
Burghardt, K.; Alsina, E. F.; Girvan, M.; Rand, W.; and Lerman, K.
PLOS ONE, 12(3): e0173610+. March 2017.
Paper
doi
link
bibtex
abstract
5 downloads
@article{Burghardt2017myopia,
abstract = {Crowds can often make better decisions than individuals or small groups of experts by leveraging their ability to aggregate diverse information. Question answering sites, such as Stack Exchange, rely on the � wisdom of crowds� effect to identify the best answers to questions asked by users. We analyze data from 250 communities on the Stack Exchange network to pinpoint factors affecting which answers are chosen as the best answers. Our results suggest that, rather than evaluate all available answers to a question, users rely on simple cognitive heuristics to choose an answer to vote for or accept. These cognitive heuristics are linked to an answer's salience, such as the order in which it is listed and how much screen space it occupies. While askers appear to depend on heuristics to a greater extent than voters when choosing an answer to accept as the most helpful one, voters use acceptance itself as a heuristic, and they are more likely to choose the answer after it has been accepted than before that answer was accepted. These heuristics become more important in explaining and predicting behavior as the number of available answers to a question increases. Our findings suggest that crowd judgments may become less reliable as the number of answers grows.},
author = {Burghardt, Keith and Alsina, Emanuel F. and Girvan, Michelle and Rand, William and Lerman, Kristina},
citeulike-article-id = {14312610},
citeulike-linkout-0 = {http://dx.doi.org/10.1371/journal.pone.0173610},
day = {16},
doi = {10.1371/journal.pone.0173610},
journal = {PLOS ONE},
keywords = {cognitive-constraints, cognitive-load, lerman, question-answering, wisdom-of-crowds},
month = mar,
number = {3},
pages = {e0173610+},
posted-at = {2017-03-17 00:37:42},
priority = {2},
publisher = {Public Library of Science},
title = {The myopia of crowds: Cognitive load and collective evaluation of answers on Stack Exchange},
url = {http://dx.doi.org/10.1371/journal.pone.0173610},
volume = {12},
year = {2017}
}
Crowds can often make better decisions than individuals or small groups of experts by leveraging their ability to aggregate diverse information. Question answering sites, such as Stack Exchange, rely on the � wisdom of crowds� effect to identify the best answers to questions asked by users. We analyze data from 250 communities on the Stack Exchange network to pinpoint factors affecting which answers are chosen as the best answers. Our results suggest that, rather than evaluate all available answers to a question, users rely on simple cognitive heuristics to choose an answer to vote for or accept. These cognitive heuristics are linked to an answer's salience, such as the order in which it is listed and how much screen space it occupies. While askers appear to depend on heuristics to a greater extent than voters when choosing an answer to accept as the most helpful one, voters use acceptance itself as a heuristic, and they are more likely to choose the answer after it has been accepted than before that answer was accepted. These heuristics become more important in explaining and predicting behavior as the number of available answers to a question increases. Our findings suggest that crowd judgments may become less reliable as the number of answers grows.
Taming the Unpredictability of Cultural Markets with Social Influence.
Abeliuk, A.; Berbeglia, G.; Hentenryck, P. V.; Hogg, T.; and Lerman, K.
In
Proceedings of the 26th International World Wide Web Conference (WWW2017), 2017.
link
bibtex
abstract
@INPROCEEDINGS{Abeliuk2017www,
author = {Andr�s Abeliuk and Gerardo Berbeglia and Pascal Van Hentenryck and Tad Hogg and Kristina Lerman},
title = {Taming the Unpredictability of Cultural Markets with Social Influence},
booktitle = {Proceedings of the 26th International World Wide Web Conference (WWW2017)},
year = {2017},
pages = {},
abstract = {Unpredictability is often portrayed as an undesirable outcome of social in uence in cultural markets. Unpredictability stems from the ``rich get richer'' effect, whereby small fluctuations in the market share or popularity of products are amplified over time by social in uence. In this paper, we report results of an experimental study that shows that unpredictability is not an inherent property of social influence. We investigate strategies for creating markets in which the popularity of products is better aligned with their underlying quality. For our study, we created a cultural market of science stories and conducted randomized experiments on different policies for presenting the stories to study participants. Speci cally, we varied how the stories were ranked, and whether or not participants were shown the ratings these stories received from others. We present a policy that leverages social influence and product positioning to help distinguish the product's market share (popularity) from underlying quality. Highlighting products with the highest estimated quality reduces the ``rich get richer'' effect of using popularity directly. We show that this policy allows us to more robustly and predictably identify high quality products and promote blockbusters. The policy can be used to create more efficient online cultural markets with a better allocation of resources to products.},
keywords = {social-dynamics},
}
Unpredictability is often portrayed as an undesirable outcome of social in uence in cultural markets. Unpredictability stems from the ``rich get richer'' effect, whereby small fluctuations in the market share or popularity of products are amplified over time by social in uence. In this paper, we report results of an experimental study that shows that unpredictability is not an inherent property of social influence. We investigate strategies for creating markets in which the popularity of products is better aligned with their underlying quality. For our study, we created a cultural market of science stories and conducted randomized experiments on different policies for presenting the stories to study participants. Speci cally, we varied how the stories were ranked, and whether or not participants were shown the ratings these stories received from others. We present a policy that leverages social influence and product positioning to help distinguish the product's market share (popularity) from underlying quality. Highlighting products with the highest estimated quality reduces the ``rich get richer'' effect of using popularity directly. We show that this policy allows us to more robustly and predictably identify high quality products and promote blockbusters. The policy can be used to create more efficient online cultural markets with a better allocation of resources to products.
iPhone's Digital Marketplace: Characterizing the Big Spenders.
Kooti, F.; Grbovic, M.; Aiello, L. M.; Bax, E.; and Lerman, K.
In
Proceedings of the 10th International ACM Conference on Web Search and Data Mining, 2017. ACM
link
bibtex
abstract
@INPROCEEDINGS{Kooti2017wsdm,
author = {Farshad Kooti and Mihajlo Grbovic and Luca Maria Aiello and Eric Bax and Kristina Lerman},
title = {iPhone's Digital Marketplace: Characterizing the Big Spenders},
booktitle = {Proceedings of the 10th International ACM Conference on Web Search and Data Mining},
year = {2017},
pages = {},
publisher = {ACM},
abstract = {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},
keywords = {social-dynamics},
}
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
Analyzing the Ride-sharing Economy.
Kooti, F.; Grbovic, M.; Aiello, L. M.; Djuric, N.; Radosavljevic, V.; and Lerman, K.
In
Proceedings of the 26th International World Wide Web Conference (Companion WWW2017), 2017.
link
bibtex
abstract
@INPROCEEDINGS{KootiG2017www,
author = {Farshad Kooti and Mihajlo Grbovic and Luca Maria Aiello and Nemanja Djuric and Vladan Radosavljevic and Kristina Lerman},
title = {Analyzing the Ride-sharing Economy},
booktitle = {Proceedings of the 26th International World Wide Web Conference (Companion WWW2017)},
year = {2017},
pages = {},
publisher = {},
abstract = {Uber is a popular ride-sharing application that matches people who need a ride with others who are willing to provide it using their personal vehicles. Uber�s success has fueled the growth of the sharing economy, where consumers and providers exchange services in a peer-to-peer fashion. Despite its growing popularity, few largescale studies examined Uber specifically, or the factors affecting user participation in the sharing economy in general. We address this gap through a large-scale study of the Uber market that analyzes 59M rides spanning a period of 7 months. These data were extracted from email receipts sent by Uber. Our data set allows us to examine the role of demographics, including age, gender, and race, on participation in the ride-sharing economy. The data is also fine-grained enough to 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 a higher driver ratings. Finally, we focus on factors that affect retention and use information from early rides to accurately predict which riders or drivers will become active Uber users.},
keywords = {social-dynamics},
}
Uber is a popular ride-sharing application that matches people who need a ride with others who are willing to provide it using their personal vehicles. Uber�s success has fueled the growth of the sharing economy, where consumers and providers exchange services in a peer-to-peer fashion. Despite its growing popularity, few largescale studies examined Uber specifically, or the factors affecting user participation in the sharing economy in general. We address this gap through a large-scale study of the Uber market that analyzes 59M rides spanning a period of 7 months. These data were extracted from email receipts sent by Uber. Our data set allows us to examine the role of demographics, including age, gender, and race, on participation in the ride-sharing economy. The data is also fine-grained enough to 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 a higher driver ratings. Finally, we focus on factors that affect retention and use information from early rides to accurately predict which riders or drivers will become active Uber users.
Understanding Short-term Changes in Online Activity Sessions.
Kooti, F.; Subbian, K.; Mason, W.; Adamic, L.; and Lerman, K.
In
Proceedings of the 26th International World Wide Web Conference (Companion WWW2017), 2017.
link
bibtex
abstract
@INPROCEEDINGS{KootiA2017www,
author = {Farshad Kooti and Karthik Subbian and Winter Mason and Lada Adamic and Kristina Lerman},
title = {Understanding Short-term Changes in Online Activity Sessions},
booktitle = {Proceedings of the 26th International World Wide Web Conference (Companion WWW2017)},
year = {2017},
pages = {},
publisher = {},
abstract = {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 person�s 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 person�s 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.},
keywords = {social-dynamics, cognitive-depletion},
}
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 person�s 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 person�s 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.
Bounded rationality in scholarly knowledge discovery.
Lerman, K.; Hodas, N.; and Wu, H.
arXiv preprint arXiv:1710.00269. 2017.
link
bibtex
@article{lerman2017bounded,
title={Bounded rationality in scholarly knowledge discovery},
author={Lerman, Kristina and Hodas, Nathan and Wu, Hao},
journal={arXiv preprint arXiv:1710.00269},
year={2017}
}
![Analyzing the Digital Traces of Political Manipulation: The 2016 Russian Interference Twitter Campaign [link]](http://bibbase.org/img/filetypes/link.svg "Paper")
Paper
link
bibtex
abstract
2 downloads
![Leveraging the Contributions of the Casual Majority to Identify AppealingWeb Content [pdf]](http://bibbase.org/img/filetypes/pdf.svg "Paper")