Jennifer Golbeck
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Jennifer Golbeck has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Aaron Mannes
University of Maryland, College Park, 8400 Baltimore Ave., College Park, MD 20740. 1.301.314.6607
2 shared publications - Cody Buntain
University of Maryland
2 shared publications - Behjat Siddiquie
Computer Vision Lab, Dept. of Computer Science, Univ. of Maryland - College Park, College Park, MD 20742, U.S.A.
1 shared publication - Irene Eleta
ISGlobal Barcelona Institute for Global Health, (CIBERESP) & Universitat Pompeu Fabra, Barcelona, Spain
1 shared publication James Hendler
MIND Lab, 8400 Baltimore Ave., Suite 200, College Park, MD 20740, U.S.A.
1 shared publication- Jennifer Golbeck
College of Information, University of Maryland, College Park, Maryland, USA
1 shared publication
Latest Publications
- 11 February 2022
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 375 (6581) , 624
Social media and shared reality Off the Edge: Flat Earthers, Conspiracy Culture, and Why People Will Believe Anything Kelly Weill Algonquin Books, 2022. 256 pp.
Exacerbated by engagement algorithms, flat Earth theories and other fringe beliefs thrive
- 11 February 2022
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 375 (6581) , 624
Social media and shared reality Off the Edge: Flat Earthers, Conspiracy Culture, and Why People Will Believe Anything Kelly Weill Algonquin Books, 2022. 256 pp.
Exacerbated by engagement algorithms, flat Earth theories and other fringe beliefs thrive
- Open Access
Automatically Identifying Fake News in Popular Twitter Threads
Information quality in social media is an increasingly important issue, but web-scale data hinders experts' ability to assess and correct much of the inaccurate content, or `fake news,' present in these platforms. This paper develops a method for automating fake news detection on Twitter by learning to predict accuracy assessments in two credibility-focused Twitter datasets: CREDBANK, a crowdsourced dataset of accuracy assessments for events in Twitter, and PHEME, a dataset of potential rumors in Twitter and journalistic assessments of their accuracies. We apply this method to Twitter content sourced from BuzzFeed's fake news dataset and show models trained against crowdsourced workers outperform models based on journalists' assessment and models trained on a pooled dataset of both crowdsourced workers and journalists. All three datasets, aligned into a uniform format, are also publicly available. A feature analysis then identifies features that are most predictive for crowdsourced and journalistic accuracy assessments, results of which are consistent with prior work. We close with a discussion contrasting accuracy and credibility and why models of non-experts outperform models of journalists for fake news detection in Twitter.
Open AccessAutomatically Identifying Fake News in Popular Twitter Threads
Information quality in social media is an increasingly important issue, but web-scale data hinders experts' ability to assess and correct much of the inaccurate content, or `fake news,' present in these platforms. This paper develops a method for automating fake news detection on Twitter by learning to predict accuracy assessments in two credibility-focused Twitter datasets: CREDBANK, a crowdsourced dataset of accuracy assessments for events in Twitter, and PHEME, a dataset of potential rumors in Twitter and journalistic assessments of their accuracies. We apply this method to Twitter content sourced from BuzzFeed's fake news dataset and show models trained against crowdsourced workers outperform models based on journalists' assessment and models trained on a pooled dataset of both crowdsourced workers and journalists. All three datasets, aligned into a uniform format, are also publicly available. A feature analysis then identifies features that are most predictive for crowdsourced and journalistic accuracy assessments, results of which are consistent with prior work. We close with a discussion contrasting accuracy and credibility and why models of non-experts outperform models of journalists for fake news detection in Twitter.
- 24 February 2017
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 355 (6327) , 804
Podcast Cannibalism A Perfectly Natural History Bill Schutt Algonquin Books, 2017. 352 pp. Homo Deus A Brief History of Tomorrow Yuval Noah Harari Harper, 2017. 455 pp.
![][1] Among invertebrates, including most predatory species of mantises, females are more likely to cannibalize than males. PHOTO: FRANCESCO TOMASINELLI/SCIENCE SOURCE This week on the Science podcast, Jennifer Golbeck explores the taboo topic of cannibalism with zoologist Bill
- 24 February 2017
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 355 (6327) , 804
Podcast Cannibalism A Perfectly Natural History Bill Schutt Algonquin Books, 2017. 352 pp. Homo Deus A Brief History of Tomorrow Yuval Noah Harari Harper, 2017. 455 pp.
![][1] Among invertebrates, including most predatory species of mantises, females are more likely to cannibalize than males. PHOTO: FRANCESCO TOMASINELLI/SCIENCE SOURCE This week on the Science podcast, Jennifer Golbeck explores the taboo topic of cannibalism with zoologist Bill
- Open Access
- 25 March 2016
- Published by MDPI AG in Future Internet
- Vol. 8 (2) , 9
User Perception of Facebook App Data Access: A Comparison of Methods and Privacy Concerns
Users share vast amounts of personal information online, but are they fully aware of what information they are sharing and with whom? In this paper, we focused on Facebook apps and set out to understand how concerned users are about privacy and how well-informed they are about what personal data apps can access. We found that initially, subjects were generally under-informed about what data apps could access from their profiles. After viewing additional information about these permissions, subjects’ concern about privacy on Facebook increased. Subjects’ understanding of what data apps were able to access increased, although even after receiving explicit information on the topic, many subjects still did not fully understand the extent to which apps could access their data.
- 25 March 2016
- Published by MDPI AG in Future Internet
- Vol. 8 (2) , 9
Open AccessUser Perception of Facebook App Data Access: A Comparison of Methods and Privacy Concerns
Users share vast amounts of personal information online, but are they fully aware of what information they are sharing and with whom? In this paper, we focused on Facebook apps and set out to understand how concerned users are about privacy and how well-informed they are about what personal data apps can access. We found that initially, subjects were generally under-informed about what data apps could access from their profiles. After viewing additional information about these permissions, subjects’ concern about privacy on Facebook increased. Subjects’ understanding of what data apps were able to access increased, although even after receiving explicit information on the topic, many subjects still did not fully understand the extent to which apps could access their data.
- Open Access
Learning to Discover Key Moments in Social Media Streams
This paper introduces LABurst, a general technique for identifying key moments, or moments of high impact, in social media streams without the need for domain-specific information or seed keywords. We leverage machine learning to model temporal patterns around bursts in Twitter's unfiltered public sample stream and build a classifier to identify tokens experiencing these bursts. We show LABurst performs competitively with existing burst detection techniques while simultaneously providing insight into and detection of unanticipated moments. To demonstrate our approach's potential, we compare two baseline event-detection algorithms with our language-agnostic algorithm to detect key moments across three major sporting competitions: 2013 World Series, 2014 Super Bowl, and 2014 World Cup. Our results show LABurst outperforms a time series analysis baseline and is competitive with a domain-specific baseline even though we operate without any domain knowledge. We then go further by transferring LABurst's models learned in the sports domain to the task of identifying earthquakes in Japan and show our method detects large spikes in earthquake-related tokens within two minutes of the actual event.
Open AccessLearning to Discover Key Moments in Social Media Streams
This paper introduces LABurst, a general technique for identifying key moments, or moments of high impact, in social media streams without the need for domain-specific information or seed keywords. We leverage machine learning to model temporal patterns around bursts in Twitter's unfiltered public sample stream and build a classifier to identify tokens experiencing these bursts. We show LABurst performs competitively with existing burst detection techniques while simultaneously providing insight into and detection of unanticipated moments. To demonstrate our approach's potential, we compare two baseline event-detection algorithms with our language-agnostic algorithm to detect key moments across three major sporting competitions: 2013 World Series, 2014 Super Bowl, and 2014 World Cup. Our results show LABurst outperforms a time series analysis baseline and is competitive with a domain-specific baseline even though we operate without any domain knowledge. We then go further by transferring LABurst's models learned in the sports domain to the task of identifying earthquakes in Japan and show our method detects large spikes in earthquake-related tokens within two minutes of the actual event.
- 8 February 2014
- Published by Springer Nature in Social Network Analysis and Mining
- Vol. 4 (1) , 1-20
Predicting personality with social behavior: a comparative study
- 8 February 2014
- Published by Springer Nature in Social Network Analysis and Mining
- Vol. 4 (1) , 1-20
Predicting personality with social behavior: a comparative study
- Open Access
- 24 January 2012
- Published by Wiley in Proceedings of the American Society for Information Science and Technology
- Vol. 49 (1) , 1-4
Bridging languages in social networks: How multilingual users of Twitter connect language communities?
Multilingual users of social media and social networking sites are invisible —even unconscious— translators. Their language skills and multicultural background could help diminish the segmentation of information spheres online by connecting different language communities. Focusing on the microblogging site Twitter, this study aims at defining and classifying language bridges by looking at the languages and structure of the social networks of multilingual individuals. The social network analysis unveils four types of structures, one of which is defined as a language bridge. The next step of this work in progress aims at providing quantitative measures of this phenomenon, which has not been yet properly described or understood.
- 24 January 2012
- Published by Wiley in Proceedings of the American Society for Information Science and Technology
- Vol. 49 (1) , 1-4
Open AccessBridging languages in social networks: How multilingual users of Twitter connect language communities?
Multilingual users of social media and social networking sites are invisible —even unconscious— translators. Their language skills and multicultural background could help diminish the segmentation of information spheres online by connecting different language communities. Focusing on the microblogging site Twitter, this study aims at defining and classifying language bridges by looking at the languages and structure of the social networks of multilingual individuals. The social network analysis unveils four types of structures, one of which is defined as a language bridge. The next step of this work in progress aims at providing quantitative measures of this phenomenon, which has not been yet properly described or understood.
- Open Access
- 20 August 2007
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 20 (5) , 431-439
A Semantic Web approach to the provenance challenge
Provenance is critically important for scientific workflow systems, as it allows users to verify data, repeat experiments, and discover dependencies. The Semantic Web is a natural fit for representing provenance, as it contains explicit support for representing and inferring connections between data and processes, as well as for adding annotations to data. In this article, we present a Semantic Web approach to the Provenance Challenge (Concurrency Computat.: Pract. Exper. 2007; DOI: 10.1002/cpe.1233). We use web services, ontologies, OWL reasoners, triple stores, and the SPARQL query language to implement the workflow, represent the data and the connections within it, and execute queries. We successfully implemented and answered all of the challenge queries. The flexibility of the Semantic Web also makes it quite easy to convert different provenance systems' data representation to a form we can work with. We illustrate this by integrating data from the PASS approach into our system, and successfully executing all of the challenge queries on it as well. Copyright © 2007 John Wiley & Sons, Ltd.
- 20 August 2007
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 20 (5) , 431-439
Open AccessA Semantic Web approach to the provenance challenge
Provenance is critically important for scientific workflow systems, as it allows users to verify data, repeat experiments, and discover dependencies. The Semantic Web is a natural fit for representing provenance, as it contains explicit support for representing and inferring connections between data and processes, as well as for adding annotations to data. In this article, we present a Semantic Web approach to the Provenance Challenge (Concurrency Computat.: Pract. Exper. 2007; DOI: 10.1002/cpe.1233). We use web services, ontologies, OWL reasoners, triple stores, and the SPARQL query language to implement the workflow, represent the data and the connections within it, and execute queries. We successfully implemented and answered all of the challenge queries. The flexibility of the Semantic Web also makes it quite easy to convert different provenance systems' data representation to a form we can work with. We illustrate this by integrating data from the PASS approach into our system, and successfully executing all of the challenge queries on it as well. Copyright © 2007 John Wiley & Sons, Ltd.
- 1 June 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 19493983,p. 85-92
Classifying Computer Generated Charts
We present an approach for classifying images of charts based on the shape and spatial relationships of their primitives. Five categories are considered: bar-charts, curve-plots, pie-charts, scatter-plots and surface-plots. We introduce two novel features to represent the structural information based on (a) region segmentation and (b) curve saliency. The local shape is characterized using the Histograms of Oriented Gradients (HOG) and the Scale Invariant Feature Transform (SIFT) descriptors. Each image is represented by sets of feature vectors of each modality. The similarity between two images is measured by the overlap in the distribution of the features -measured using the Pyramid Match algorithm. A test image is classified based on its similarity with training images from the categories. The approach is tested with a database of images collected from the Internet.
- 1 June 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 19493983,p. 85-92
Classifying Computer Generated Charts
We present an approach for classifying images of charts based on the shape and spatial relationships of their primitives. Five categories are considered: bar-charts, curve-plots, pie-charts, scatter-plots and surface-plots. We introduce two novel features to represent the structural information based on (a) region segmentation and (b) curve saliency. The local shape is characterized using the Histograms of Oriented Gradients (HOG) and the Scale Invariant Feature Transform (SIFT) descriptors. Each image is represented by sets of feature vectors of each modality. The similarity between two images is measured by the overlap in the distribution of the features -measured using the Pyramid Match algorithm. A test image is classified based on its similarity with training images from the categories. The approach is tested with a database of images collected from the Internet.
- 1 January 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Ontology Building: A Terrorism Specialist's Perspective
For the semantic Web to fulfill its potential, it needs to be usable by the average user. Until this occurs, the semantic Web will remain the province of programmers and computer scientists. Ontologies, which define concepts for the computer, give the semantic Web its potential power. However, until non-computer science specialists are able to construct useful ontologies, the semantic Web will not fulfill its potential. This paper discusses a non-computer scientist's efforts to construct an ontology for a terrorism analysis Semantic Web portal (profilesinterror.mindswap.org). The paper will address both general challenges in conceptualizing an ontology and specific techniques useful for ontology construction. Some of the concepts will be useful for ontology writing in general, other will be unique to defining terrorism.
- 1 January 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Ontology Building: A Terrorism Specialist's Perspective
For the semantic Web to fulfill its potential, it needs to be usable by the average user. Until this occurs, the semantic Web will remain the province of programmers and computer scientists. Ontologies, which define concepts for the computer, give the semantic Web its potential power. However, until non-computer science specialists are able to construct useful ontologies, the semantic Web will not fulfill its potential. This paper discusses a non-computer scientist's efforts to construct an ontology for a terrorism analysis Semantic Web portal (profilesinterror.mindswap.org). The paper will address both general challenges in conceptualizing an ontology and specific techniques useful for ontology construction. Some of the concepts will be useful for ontology writing in general, other will be unique to defining terrorism.