Search

Child pornography represents a severe form of exploitation and victimization of children, leaving the victims with emotional and physical trauma. In this study, we aim to analyze local patterns of child pornography consumption across 1341 French communes in 20 metropolitan regions of France using fine-grained mobile traffic data of Tor network-related web services. We estimate that approx. 0.08 % of Tor mobile download traffic observed in France is linked to the consumption of child sexual abuse materials by correlating it with local-level temporal porn consumption patterns. This compares to 0.19 % of what we conservatively estimate to be the share of child pornographic content in global Tor traffic. In line with existing literature on the link between sexual child abuse and the consumption of image-based content thereof, we observe a positive and statistically significant effect of our child pornography consumption estimates on the reported number of victims of sexual violence and vice versa, which validates our findings, after controlling for a set of spatial and non-spatial features including socio-demographic characteristics, voting behaviour, nearby points of interest and Google Trends queries. While this is a first, exploratory attempt to look at child pornography from a spatial epidemiological angle, we believe this research provides public health officials with valuable information to prioritize target areas for public awareness campaigns as another step to fulfil the global community's pledge to target 16.2 of the Sustainable Development Goals: "End abuse, exploitation, trafficking and all forms of violence and torture against children"., Comment: In the first version of the arXiv preprint of this paper, we reported this share to be 16.9 %. This was based on a misinterpretation of the Tor statistics. After expert discussions, we corrected it and any subsequent analysis

Online professional social networking platforms provide opportunities to expand networks strategically for job opportunities and career advancement. A large body of research shows that women's offline networks are less advantageous than men's. How online platforms such as LinkedIn may reflect or reproduce gendered networking behaviours, or how online social connectivity may affect outcomes differentially by gender is not well understood. This paper analyses aggregate, anonymised data from almost 10 million LinkedIn users in the UK and US information technology (IT) sector collected from the site's advertising platform to explore how being connected to Big Tech companies ('social connectivity') varies by gender, and how gender, age, seniority and social connectivity shape the propensity to report job promotions or relocations. Consistent with previous studies, we find there are fewer women compared to men on LinkedIn in IT. Furthermore, female users are less likely to be connected to Big Tech companies than men. However, when we further analyse recent promotion or relocation reports, we find women are more likely than men to have reported a recent promotion at work, suggesting high-achieving women may be self-selecting onto LinkedIn. Even among this positively selected group, though, we find men are more likely to report a recent relocation. Social connectivity emerges as a significant predictor of promotion and relocation reports, with an interaction effect between gender and social connectivity indicating the payoffs to social connectivity for promotion and relocation reports are larger for women. This suggests that online networking has the potential for larger impacts for women, who experience greater disadvantage in traditional networking contexts, and calls for further research to understand differential impacts of online networking for socially disadvantaged groups., Comment: Accepted and forthcoming at the International AAAI Conference on Web and Social Media (ICWSM) 2024

It is ten years since neural networks made their spectacular comeback. Prompted by this anniversary, we take a holistic perspective on Artificial Intelligence (AI). Supervised Learning for cognitive tasks is effectively solved - provided we have enough high-quality labeled data. However, deep neural network models are not easily interpretable, and thus the debate between blackbox and whitebox modeling has come to the fore. The rise of attention networks, self-supervised learning, generative modeling, and graph neural networks has widened the application space of AI. Deep Learning has also propelled the return of reinforcement learning as a core building block of autonomous decision making systems. The possible harms made possible by new AI technologies have raised socio-technical issues such as transparency, fairness, and accountability. The dominance of AI by Big-Tech who control talent, computing resources, and most importantly, data may lead to an extreme AI divide. Failure to meet high expectations in high profile, and much heralded flagship projects like self-driving vehicles could trigger another AI winter.

We investigate how representations of Syrian refugees (2011-2021) differ across US partisan news outlets. We analyze 47,388 articles from the online US media about Syrian refugees to detail differences in reporting between left- and right-leaning media. We use various NLP techniques to understand these differences. Our polarization and question answering results indicated that left-leaning media tended to represent refugees as child victims, welcome in the US, and right-leaning media cast refugees as Islamic terrorists. We noted similar results with our sentiment and offensive speech scores over time, which detail possibly unfavorable representations of refugees in right-leaning media. A strength of our work is how the different techniques we have applied validate each other. Based on our results, we provide several recommendations. Stakeholders may utilize our findings to intervene around refugee representations, and design communications campaigns that improve the way society sees refugees and possibly aid refugee outcomes.

Social media users post content on various topics. A defining feature of social media is that other users can provide feedback -- called community feedback -- to their content in the form of comments, replies, and retweets. We hypothesize that the amount of received feedback influences the choice of topics on which a social media user posts. However, it is challenging to test this hypothesis as user heterogeneity and external confounders complicate measuring the feedback effect. Here, we investigate this hypothesis with a predictive approach based on an interpretable model of an author's decision to continue the topic of their previous post. We explore the confounding factors, including author's topic preferences and unobserved external factors such as news and social events, by optimizing the predictive accuracy. This approach enables us to identify which users are susceptible to community feedback. Overall, we find that 33\% and 14\% of active users in Reddit and Twitter, respectively, are influenced by community feedback. The model suggests that this feedback alters the probability of topic continuation up to 14\%, depending on the user and the amount of feedback., Comment: Published at EPJ Data science in 2020

In the global move toward urbanization, making sure the people remaining in rural areas are not left behind in terms of development and policy considerations is a priority for governments worldwide. However, it is increasingly challenging to track important statistics concerning this sparse, geographically dispersed population, resulting in a lack of reliable, up-to-date data. In this study, we examine the usefulness of the Facebook Advertising platform, which offers a digital "census" of over two billions of its users, in measuring potential rural-urban inequalities. We focus on Italy, a country where about 30% of the population lives in rural areas. First, we show that the population statistics that Facebook produces suffer from instability across time and incomplete coverage of sparsely populated municipalities. To overcome such limitation, we propose an alternative methodology for estimating Facebook Ads audiences that nearly triples the coverage of the rural municipalities from 19% to 55% and makes feasible fine-grained sub-population analysis. Using official national census data, we evaluate our approach and confirm known significant urban-rural divides in terms of educational attainment and income. Extending the analysis to Facebook-specific user "interests" and behaviors, we provide further insights on the divide, for instance, finding that rural areas show a higher interest in gambling. Notably, we find that the most predictive features of income in rural areas differ from those for urban centres, suggesting researchers need to consider a broader range of attributes when examining rural wellbeing. The findings of this study illustrate the necessity of improving existing tools and methodologies to include under-represented populations in digital demographic studies -- the failure to do so could result in misleading observations, conclusions, and most importantly, policies., Comment: To be published in the Proceedings of The Web Conference 2020 (WWW '20)

In online social networks, it is common to use predictions of node categories to estimate measures of homophily and other relational properties. However, online social network data often lacks basic demographic information about the nodes. Researchers must rely on predicted node attributes to estimate measures of homophily, but little is known about the validity of these measures. We show that estimating homophily in a network can be viewed as a dyadic prediction problem, and that homophily estimates are unbiased when dyad-level residuals sum to zero in the network. Node-level prediction models, such as the use of names to classify ethnicity or gender, do not generally have this property and can introduce large biases into homophily estimates. Bias occurs due to error autocorrelation along dyads. Importantly, node-level classification performance is not a reliable indicator of estimation accuracy for homophily. We compare estimation strategies that make predictions at the node and dyad levels, evaluating performance in different settings. We propose a novel "ego-alter" modeling approach that outperforms standard node and dyad classification strategies. While this paper focuses on homophily, results generalize to other relational measures which aggregate predictions along the dyads in a network. We conclude with suggestions for research designs to study homophily in online networks. Code for this paper is available at https://github.com/georgeberry/autocorr., Comment: 19 pages, 4 figures, 2 tables

A food recipe is an ordered set of instructions for preparing a particular dish. From a visual perspective, every instruction step can be seen as a way to change the visual appearance of the dish by adding extra objects (e.g., adding an ingredient) or changing the appearance of the existing ones (e.g., cooking the dish). In this paper, we aim to teach a machine how to make a pizza by building a generative model that mirrors this step-by-step procedure. To do so, we learn composable module operations which are able to either add or remove a particular ingredient. Each operator is designed as a Generative Adversarial Network (GAN). Given only weak image-level supervision, the operators are trained to generate a visual layer that needs to be added to or removed from the existing image. The proposed model is able to decompose an image into an ordered sequence of layers by applying sequentially in the right order the corresponding removing modules. Experimental results on synthetic and real pizza images demonstrate that our proposed model is able to: (1) segment pizza toppings in a weaklysupervised fashion, (2) remove them by revealing what is occluded underneath them (i.e., inpainting), and (3) infer the ordering of the toppings without any depth ordering supervision. Code, data, and models are available online., Comment: CVPR 2019

Technologies for abusive language detection are being developed and applied with little consideration of their potential biases. We examine racial bias in five different sets of Twitter data annotated for hate speech and abusive language. We train classifiers on these datasets and compare the predictions of these classifiers on tweets written in African-American English with those written in Standard American English. The results show evidence of systematic racial bias in all datasets, as classifiers trained on them tend to predict that tweets written in African-American English are abusive at substantially higher rates. If these abusive language detection systems are used in the field they will therefore have a disproportionate negative impact on African-American social media users. Consequently, these systems may discriminate against the groups who are often the targets of the abuse we are trying to detect., Comment: To appear in the proceedings of the Third Abusive Language Workshop (https://sites.google.com/view/alw3/) at the Annual Meeting for the Association for Computational Linguistics 2019. Please cite the published version

The degree to which Mexican immigrants in the U.S. are assimilating culturally has been widely debated. To examine this question, we focus on musical taste, a key symbolic resource that signals the social positions of individuals. We adapt an assimilation metric from earlier work to analyze self-reported musical interests among immigrants in Facebook. We use the relative levels of interest in musical genres, where a similarity to the host population in musical preferences is treated as evidence of cultural assimilation. Contrary to skeptics of Mexican assimilation, we find significant cultural convergence even among first-generation immigrants, which problematizes their use as assimilative "benchmarks" in the literature. Further, 2nd generation Mexican Americans show high cultural convergence vis-\`a-vis both Anglos and African-Americans, with the exception of those who speak Spanish. Rather than conforming to a single assimilation path, our findings reveal how Mexican immigrants defy simple unilinear theoretical expectations and illuminate their uniquely heterogeneous character., Comment: WebConf 2019

In this paper, we introduce Recipe1M+, a new large-scale, structured corpus of over one million cooking recipes and 13 million food images. As the largest publicly available collection of recipe data, Recipe1M+ affords the ability to train high-capacity modelson aligned, multimodal data. Using these data, we train a neural network to learn a joint embedding of recipes and images that yields impressive results on an image-recipe retrieval task. Moreover, we demonstrate that regularization via the addition of a high-level classification objective both improves retrieval performance to rival that of humans and enables semantic vector arithmetic. We postulate that these embeddings will provide a basis for further exploration of the Recipe1M+ dataset and food and cooking in general. Code, data and models are publicly available., Comment: IEEE Transactions on Pattern Analysis and Machine Intelligence

We consider the problem of obtaining unbiased estimates of group properties in social networks when using a classifier for node labels. Inference for this problem is complicated by two factors: the network is not known and must be crawled, and even high-performance classifiers provide biased estimates of group proportions. We propose and evaluate AdjustedWalk for addressing this problem. This is a three step procedure which entails: 1) walking the graph starting from an arbitrary node; 2) learning a classifier on the nodes in the walk; and 3) applying a post-hoc adjustment to classification labels. The walk step provides the information necessary to make inferences over the nodes and edges, while the adjustment step corrects for classifier bias in estimating group proportions. This process provides de-biased estimates at the cost of additional variance. We evaluate AdjustedWalk on four tasks: the proportion of nodes belonging to a minority group, the proportion of the minority group among high degree nodes, the proportion of within-group edges, and Coleman's homophily index. Simulated and empirical graphs show that this procedure performs well compared to optimal baselines in a variety of circumstances, while indicating that variance increases can be large for low-recall classifiers., Comment: 19 pages, 6 figures, 1 table

In many developing countries, timely and accurate information about birth rates and other demographic indicators is still lacking, especially for male fertility rates. Using anonymous and aggregate data from Facebook's Advertising Platform, we produce global estimates of the Mean Age at Childbearing (MAC), a key indicator of fertility postponement. Our analysis indicates that fertility measures based on Facebook data are highly correlated with conventional indicators based on traditional data, for those countries for which we have statistics. For instance, the correlation of the MAC computed using Facebook and United Nations data is 0.47 (p = 4.02e-08) and 0.79 (p = 2.2e-15) for female and male respectively. Out of sample validation for a simple regression model indicates that the mean absolute percentage error is 2.3%. We use the linear model and Facebook data to produce estimates of the male MAC for countries for which we do not have data., Comment: Please cite the version from Proceedings of the Twelfth International Conference on Web and Social Media (ICWSM-2018)

Gender imbalances in work environments have been a long-standing concern. Identifying the existence of such imbalances is key to designing policies to help overcome them. In this work, we study gender trends in employment across various dimensions in the United States. This is done by analyzing anonymous, aggregate statistics that were extracted from LinkedIn's advertising platform. The data contain the number of male and female LinkedIn users with respect to (i) location, (ii) age, (iii) industry and (iv) certain skills. We studied which of these categories correlate the most with high relative male or female presence on LinkedIn. In addition to examining the summary statistics of the LinkedIn data, we model the gender balance as a function of the different employee features using linear regression. Our results suggest that the gender gap varies across all feature types, but the differences are most profound among industries and skills. A high correlation between gender ratios of people in our LinkedIn data set and data provided by the US Bureau of Labor Statistics serves as external validation for our results., Comment: Accepted at a poster at ICWSM 2018. Please cite the ICWSM version

Migrants' assimilation is a major challenge for European societies, in part because of the sudden surge of refugees in recent years and in part because of long-term demographic trends. In this paper, we use Facebook's data for advertisers to study the levels of assimilation of Arabic-speaking migrants in Germany, as seen through the interests they express online. Our results indicate a gradient of assimilation along demographic lines, language spoken and country of origin. Given the difficulty to collect timely migration data, in particular for traits related to cultural assimilation, the methods that we develop and the results that we provide open new lines of research that computational social scientists are well-positioned to address., Comment: Accepted as a short paper at Social Informatics 2018 (https://socinfo2018.hse.ru/). Please cite the SocInfo version

As the body of research on abusive language detection and analysis grows, there is a need for critical consideration of the relationships between different subtasks that have been grouped under this label. Based on work on hate speech, cyberbullying, and online abuse we propose a typology that captures central similarities and differences between subtasks and we discuss its implications for data annotation and feature construction. We emphasize the practical actions that can be taken by researchers to best approach their abusive language detection subtask of interest., Comment: To appear in the proceedings of the 1st Workshop on Abusive Language Online. Please cite that version

Celebrity and fandom have been studied extensively in real life. However, with more and more celebrities using social media, the dynamics of interaction between celebrities and fans has changed. Using data from a set of 57,000 fans for the top followed celebrities on Twitter, we define a wide range of features based on their Twitter activity. Using factor analysis we find the most important factors that underlie fan behavior. Using these factors, we conduct analysis on (i) understanding fan behavior by gender \& age, and (ii) para-social breakup behavior. We find that (i) fandom is a social phenomenon, (ii) female fans are often more devoted and younger fans are more active & social, and (iii) the most devoted fans are more likely to be involved in a para-social breakup. Our findings confirm existing research on para-social interactions. Given the scale of our study and dependence on non-reactive data, our paper opens new avenues for research in para-social interactions., Comment: Accepted as a poster at WebSci 2017. This is a longer version of the 2-page poster published at WebSci

Every day, millions of users reveal their interests on Facebook, which are then monetized via targeted advertisement marketing campaigns. In this paper, we explore the use of demographically rich Facebook Ads audience estimates for tracking non-communicable diseases around the world. Across 47 countries, we compute the audiences of marker interests, and evaluate their potential in tracking health conditions associated with tobacco use, obesity, and diabetes, compared to the performance of placebo interests. Despite its huge potential, we find that, for modeling prevalence of health conditions across countries, differences in these interest audiences are only weakly indicative of the corresponding prevalence rates. Within the countries, however, our approach provides interesting insights on trends of health awareness across demographic groups. Finally, we provide a temporal error analysis to expose the potential pitfalls of using Facebook's Marketing API as a black box., Comment: Please cite the article published at WebSci'17 instead of this arxiv version

Global recruitment into radical Islamic movements has spurred renewed interest in the appeal of political extremism. Is the appeal a rational response to material conditions or is it the expression of psychological and personality disorders associated with aggressive behavior, intolerance, conspiratorial imagination, and paranoia? Empirical answers using surveys have been limited by lack of access to extremist groups, while field studies have lacked psychological measures and failed to compare extremists with contrast groups. We revisit the debate over the appeal of extremism in the U.S. context by comparing publicly available Twitter messages written by over 355,000 political extremist followers with messages written by non-extremist U.S. users. Analysis of text-based psychological indicators supports the moral foundation theory which identifies emotion as a critical factor in determining political orientation of individuals. Extremist followers also differ from others in four of the Big Five personality traits., Comment: 11 pages, 3 figures

A key challenge for automatic hate-speech detection on social media is the separation of hate speech from other instances of offensive language. Lexical detection methods tend to have low precision because they classify all messages containing particular terms as hate speech and previous work using supervised learning has failed to distinguish between the two categories. We used a crowd-sourced hate speech lexicon to collect tweets containing hate speech keywords. We use crowd-sourcing to label a sample of these tweets into three categories: those containing hate speech, only offensive language, and those with neither. We train a multi-class classifier to distinguish between these different categories. Close analysis of the predictions and the errors shows when we can reliably separate hate speech from other offensive language and when this differentiation is more difficult. We find that racist and homophobic tweets are more likely to be classified as hate speech but that sexist tweets are generally classified as offensive. Tweets without explicit hate keywords are also more difficult to classify., Comment: To appear in the Proceedings of ICWSM 2017. Please cite that version

Fertility choices are linked to the different preferences and constraints of individuals and couples, and vary importantly by socio-economic status, as well by cultural and institutional context. The meaning of childbearing and child-rearing, therefore, differs between individuals and across groups. In this paper, we combine data from Google Correlate and Google Trends for the U.S. with ground truth data from the American Community Survey to derive new insights into fertility and its meaning. First, we show that Google Correlate can be used to illustrate socio-economic differences on the circumstances around pregnancy and birth: e.g., searches for "flying while pregnant" are linked to high income fertility, and "paternity test" are linked to non-marital fertility. Second, we combine several search queries to build predictive models of regional variation in fertility, explaining about 75% of the variance. Third, we explore if aggregated web search data can also be used to model fertility trends., Comment: This is a preprint of a short paper accepted at ICWSM'17. Please cite that version instead

A person's weight status can have profound implications on their life, ranging from mental health, to longevity, to financial income. At the societal level, "fat shaming" and other forms of "sizeism" are a growing concern, while increasing obesity rates are linked to ever raising healthcare costs. For these reasons, researchers from a variety of backgrounds are interested in studying obesity from all angles. To obtain data, traditionally, a person would have to accurately self-report their body-mass index (BMI) or would have to see a doctor to have it measured. In this paper, we show how computer vision can be used to infer a person's BMI from social media images. We hope that our tool, which we release, helps to advance the study of social aspects related to body weight., Comment: This is a preprint of a short paper accepted at ICWSM'17. Please cite that version instead

Social media has played an important role in shaping political discourse over the last decade. At the same time, it is often perceived to have increased political polarization, thanks to the scale of discussions and their public nature. In this paper, we try to answer the question of whether political polarization in the US on Twitter has increased over the last eight years. We analyze a large longitudinal Twitter dataset of 679,000 users and look at signs of polarization in their (i) network - how people follow political and media accounts, (ii) tweeting behavior - whether they retweet content from both sides, and (iii) content - how partisan the hashtags they use are. Our analysis shows that online polarization has indeed increased over the past eight years and that, depending on the measure, the relative change is 10%-20%. Our study is one of very few with such a long-term perspective, encompassing two US presidential elections and two mid-term elections, providing a rare longitudinal analysis., Comment: This is a preprint of a short paper accepted at ICWSM'17. Please cite that version instead

Food is an integral part of our life and what and how much we eat crucially affects our health. Our food choices largely depend on how we perceive certain characteristics of food, such as whether it is healthy, delicious or if it qualifies as a salad. But these perceptions differ from person to person and one person's "single lettuce leaf" might be another person's "side salad". Studying how food is perceived in relation to what it actually is typically involves a laboratory setup. Here we propose to use recent advances in image recognition to tackle this problem. Concretely, we use data for 1.9 million images from Instagram from the US to look at systematic differences in how a machine would objectively label an image compared to how a human subjectively does. We show that this difference, which we call the "perception gap", relates to a number of health outcomes observed at the county level. To the best of our knowledge, this is the first time that image recognition is being used to study the "misalignment" of how people describe food images vs. what they actually depict., Comment: This is a pre-print of our paper accepted to appear in the Proceedings of 2017 International World Wide Web Conference (WWW'17)

Social feedback has long been recognized as an important element of successful health-related behavior change. However, most of the existing studies look at the effect that offline social feedback has. This paper fills gaps in the literature by proposing a framework to study the causal effect that receiving social support in the form of comments in an online weight loss community has on (i) the probability of the user to return to the forum, and, more importantly, on (ii) the weight loss reported by the user. Using a matching approach for causal inference we observe a difference of 9 lbs lost between users who do or do not receive comments. Surprisingly, this effect is mediated by neither an increase in lifetime in the community nor by an increased activity level of the user. Our results show the importance that a "warm welcome" has when using online support forums to achieve health outcomes., Comment: This is a preprint of an article appearing at WWW 2017

In this paper, we explore the problem of identifying substitute relationship between food pairs from real-world food consumption data as the first step towards the healthier food recommendation. Our method is inspired by the distributional hypothesis in linguistics. Specifically, we assume that foods that are consumed in similar contexts are more likely to be similar dietarily. For example, a turkey sandwich can be considered a suitable substitute for a chicken sandwich if both tend to be consumed with french fries and salad. To evaluate our method, we constructed a real-world food consumption dataset from MyFitnessPal's public food diary entries and obtained ground-truth human judgements of food substitutes from a crowdsourcing service. The experiment results suggest the effectiveness of the method in identifying suitable substitutes., Comment: To appear at HealthRecSys'16

Social media enjoys a growing popularity as a platform to seek and share personal health information. For sleep studies using data from social media, most researchers focused on inferring sleep-related artifacts from self-reported anecdotal pointers to sleep patterns or issues such as insomnia. The data shared by "quantified-selfers" on social media presents an opportunity to study more quantitative and objective measures of sleep. We propose and validate the approach of collecting and analyzing sleep logs that are generated and shared through a sleep-tracking mobile application. We highlight the value of this data by combining it with users' social media data. The results provide a validation of using social media for sleep studies as the collected sleep data is aligned with sleep data from other sources. The results of combining social media data with sleep data provide preliminary evidence that higher social media activity is associated with lower sleep duration and quality., Comment: This is a preprint of an article accepted to appear at IEEE ICHI 2016

Recently, there have been considerable efforts to use online data to investigate international migration. These efforts show that Web data are valuable for estimating migration rates and are relatively easy to obtain. However, existing studies have only investigated flows of people along migration corridors, i.e. between pairs of countries. In our work, we use data about "places lived" from millions of Google+ users in order to study migration "clusters", i.e. groups of countries in which individuals have lived. For the first time, we consider information about more than two countries people have lived in. We argue that these data are very valuable because this type of information is not available in traditional demographic sources which record country-to-country migration flows independent of each other. We show that migration clusters of country triads cannot be identified using information about bilateral flows alone. To demonstrate the additional insights that can be gained by using data about migration clusters, we first develop a model that tries to predict the prevalence of a given triad using only data about its constituent pairs. We then inspect the groups of three countries which are more or less prominent, compared to what we would expect based on bilateral flows alone. Next, we identify a set of features such as a shared language or colonial ties that explain which triple of country pairs are more or less likely to be clustered when looking at country triples. Then we select and contrast a few cases of clusters that provide some qualitative information about what our data set shows. The type of data that we use is potentially available for a number of social media services. We hope that this first study about migration clusters will stimulate the use of Web data for the development of new theories of international migration that could not be tested appropriately before., Comment: 2016 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM)

Musicologists and sociologists have long been interested in patterns of music consumption and their relation to socioeconomic status. In particular, the Omnivore Thesis examines the relationship between these variables and the diversity of music a person consumes. Using data from social media users of Last.fm and Twitter, we design and evaluate a measure that reasonably captures diversity of musical tastes. We use that measure to explore associations between musical diversity and variables that capture socioeconomic status, demographics, and personal traits such as openness and degree of interest in music (into-ness). Our musical diversity measure can provide a useful means for studies of musical preferences and consumption. Also, our study of the Omnivore Thesis provides insights that extend previous survey and interview-based studies., Comment: 10 pages, 1 figure, 2 tables, Accepted to the 10th International AAAI Conference on Web and Social Media, ICWSM'16

Social media platforms provide several social interactional features. Due to the large scale reach of social media, these interactional features help enable various types of political discourse. Constructive and diversified discourse is important for sustaining healthy communities and reducing the impact of echo chambers. In this paper, we empirically examine the role of a newly introduced Twitter feature, 'quote retweets' (or 'quote RTs') in political discourse, specifically whether it has led to improved, civil, and balanced exchange. Quote RTs allow users to quote the tweet they retweet, while adding a short comment. Our analysis using content, network and crowd labeled data indicates that the feature has increased political discourse and its diffusion, compared to existing features. We discuss the implications of our findings in understanding and reducing online polarization., Comment: Accepted as short paper at ACM WebScience 2016

Social media systems allow Internet users a congenial platform to freely express their thoughts and opinions. Although this property represents incredible and unique communication opportunities, it also brings along important challenges. Online hate speech is an archetypal example of such challenges. Despite its magnitude and scale, there is a significant gap in understanding the nature of hate speech on social media. In this paper, we provide the first of a kind systematic large scale measurement study of the main targets of hate speech in online social media. To do that, we gather traces from two social media systems: Whisper and Twitter. We then develop and validate a methodology to identify hate speech on both these systems. Our results identify online hate speech forms and offer a broader understanding of the phenomenon, providing directions for prevention and detection approaches., Comment: Short paper, 4 pages, 4 tables

On social media platforms, like Twitter, users are often interested in gaining more influence and popularity by growing their set of followers, aka their audience. Several studies have described the properties of users on Twitter based on static snapshots of their follower network. Other studies have analyzed the general process of link formation. Here, rather than investigating the dynamics of this process itself, we study how the characteristics of the audience and follower links change as the audience of a user grows in size on the road to user's popularity. To begin with, we find that the early followers tend to be more elite users than the late followers, i.e., they are more likely to have verified and expert accounts. Moreover, the early followers are significantly more similar to the person that they follow than the late followers. Namely, they are more likely to share time zone, language, and topics of interests with the followed user. To some extent, these phenomena are related with the growth of Twitter itself, wherein the early followers tend to be the early adopters of Twitter, while the late followers are late adopters. We isolate, however, the effect of the growth of audiences consisting of followers from the growth of Twitter's user base itself. Finally, we measure the engagement of such audiences with the content of the followed user, by measuring the probability that an early or late follower becomes a retweeter., Comment: Accepted to ICWSM'16, 4 pages, 5 figures, In Proceedings of the 10th International AAAI Conference on Web and Social Media

Demographics, in particular, gender, age, and race, are a key predictor of human behavior. Despite the significant effect that demographics plays, most scientific studies using online social media do not consider this factor, mainly due to the lack of such information. In this work, we use state-of-the-art face analysis software to infer gender, age, and race from profile images of 350K Twitter users from New York. For the period from November 1, 2014 to October 31, 2015, we study which hashtags are used by different demographic groups. Though we find considerable overlap for the most popular hashtags, there are also many group-specific hashtags., Comment: This is a preprint of an article appearing at ICWSM 2016

Understanding the demographics of app users is crucial, for example, for app developers, who wish to target their advertisements more effectively. Our work addresses this need by studying the predictability of user demographics based on the list of a user's apps which is readily available to many app developers. We extend previous work on the problem on three frontiers: (1) We predict new demographics (age, race, and income) and analyze the most informative apps for four demographic attributes included in our analysis. The most predictable attribute is gender (82.3 % accuracy), whereas the hardest to predict is income (60.3 % accuracy). (2) We compare several dimensionality reduction methods for high-dimensional app data, finding out that an unsupervised method yields superior results compared to aggregating the apps at the app category level, but the best results are obtained simply by the raw list of apps. (3) We look into the effect of the training set size and the number of apps on the predictability and show that both of these factors have a large impact on the prediction accuracy. The predictability increases, or in other words, a user's privacy decreases, the more apps the user has used, but somewhat surprisingly, after 100 apps, the prediction accuracy starts to decrease., Comment: This is a pre-print of an article appearing at ICWSM 2016

An increasing number of people use wearables and other smart devices to quantify various health conditions, ranging from sleep patterns, to body weight, to heart rates. Of these Quantified Selfs many choose to openly share their data via online social networks such as Twitter and Facebook. In this study, we use data for users who have chosen to connect their smart scales to Twitter, providing both a reliable time series of their body weight, as well as insights into their social surroundings and general online behavior. Concretely, we look at which social media features are predictive of physical status, such as body weight at the individual level, and activity patterns at the population level. We show that it is possible to predict an individual's weight using their online social behaviors, such as their self-description and tweets. Weekly and monthly patterns of quantified-self behaviors are also discovered. These findings could contribute to building models to monitor public health and to have more customized personal training interventions. While there are many studies using either quantified self or social media data in isolation, this is one of the few that combines the two data sources and, to the best of our knowledge, the only one that uses public data.

It is generally accepted as common wisdom that receiving social feedback is helpful to (i) keep an individual engaged with a community and to (ii) facilitate an individual's positive behavior change. However, quantitative data on the effect of social feedback on continued engagement in an online health community is scarce. In this work we apply Mahalanobis Distance Matching (MDM) to demonstrate the importance of receiving feedback in the "loseit" weight loss community on Reddit. Concretely we show that (i) even when correcting for differences in word choice, users receiving more positive feedback on their initial post are more likely to return in the future, and that (ii) there are diminishing returns and social feedback on later posts is less important than for the first post. We also give a description of the type of initial posts that are more likely to attract this valuable social feedback. Though we cannot yet argue about ultimate weight loss success or failure, we believe that understanding the social dynamics underlying online health communities is an important step to devise more effective interventions., Comment: This is a preprint of an article appearing at ACM Digital Health 2016

To use social media for health-related analysis, one key step is the detection of health-related labels for users. But unlike transient conditions like flu, social media users are less vocal about chronic conditions such as obesity, as users might not tweet "I'm still overweight". As, however, obesity-related conditions such as diabetes, heart disease, osteoarthritis, and even cancer are on the rise, this obese-or-not label could be one of the most useful for studies in public health. In this paper we investigate the feasibility of using profile pictures to infer if a user is overweight or not. We show that this is indeed possible and further show that the fraction of labeled-as-overweight users is higher in U.S. counties with higher obesity rates. Going from public to individual health analysis, we then find differences both in behavior and social networks, for example finding users labeled as overweight to have fewer followers., Comment: This is a preprint of an article appearing at ACM DigitalHealth 2016

The debates on minority issues are often dominated by or held among the concerned minority: gender equality debates have often failed to engage men, while those about race fail to effectively engage the dominant group. To test this observation, we study the #BlackLivesMatter}movement and hashtag on Twitter--which has emerged and gained traction after a series of events typically involving the death of African-Americans as a result of police brutality--and aim to quantify the population biases across user types (individuals vs. organizations), and (for individuals) across various demographics factors (race, gender and age). Our results suggest that more African-Americans engage with the hashtag, and that they are also more active than other demographic groups. We also discuss ethical caveats with broader implications for studies on sensitive topics (e.g. discrimination, mental health, or religion) that focus on users., Comment: 9 pages, 7 figures, accepted to AAAI Spring Symposia on Observational Studies through Social Media and Other Human-Generated Content, Stanford, US, March 2016

Several projects have shown the feasibility to use textual social media data to track public health concerns, such as temporal influenza patterns or geographical obesity patterns. In this paper, we look at whether geo-tagged images from Instagram also provide a viable data source. Especially for "lifestyle" diseases, such as obesity, drinking or smoking, images of social gatherings could provide information that is not necessarily shared in, say, tweets. In this study, we explore whether (i) tags provided by the users and (ii) annotations obtained via automatic image tagging are indeed valuable for studying public health. We find that both user-provided and machine-generated tags provide information that can be used to infer a county's health statistics. Whereas for most statistics user-provided tags are better features, for predicting excessive drinking machine-generated tags such as "liquid" and "glass" yield better models. This hints at the potential of using machine-generated tags to study substance abuse., Comment: Accepted at CHI 2016 as a note

Given the ever increasing amount of publicly available social media data, there is growing interest in using online data to study and quantify phenomena in the offline "real" world. As social media data can be obtained in near real-time and at low cost, it is often used for "now-casting" indices such as levels of flu activity or unemployment. The term "social sensing" is often used in this context to describe the idea that users act as "sensors", publicly reporting their health status or job losses. Sensor activity during a time period is then typically aggregated in a "one tweet, one vote" fashion by simply counting. At the same time, researchers readily admit that social media users are not a perfect representation of the actual population. Additionally, users differ in the amount of details of their personal lives that they reveal. Intuitively, it should be possible to improve now-casting by assigning different weights to different user groups. In this paper, we ask "How does social sensing actually work?" or, more precisely, "Whom should we sense--and whom not--for optimal results?". We investigate how different sampling strategies affect the performance of now-casting of two common offline indices: flu activity and unemployment rate. We show that now-casting can be improved by 1) applying user filtering techniques and 2) selecting users with complete profiles. We also find that, using the right type of user groups, now-casting performance does not degrade, even when drastically reducing the size of the dataset. More fundamentally, we describe which type of users contribute most to the accuracy by asking if "babblers are better". We conclude the paper by providing guidance on how to select better user groups for more accurate now-casting., Comment: This is a pre-print of a forthcoming EPJ Data Science paper