Fabio Petroni

The spreading of misinformation online

Significance The wide availability of user-provided content in online social media facilitates the aggregation of people around common interests, worldviews, and narratives. However, the World Wide Web is a fruitful environment for the massive diffusion of unverified rumors. In this work, using a massive quantitative analysis of Facebook, we show that information related to distinct narratives––conspiracy theories and scientific news––generates homogeneous and polarized communities (i.e., echo chambers) having similar information consumption patterns. Then, we derive a data-driven percolation model of rumor spreading that demonstrates that homogeneity and polarization are the main determinants for predicting cascades’ size. The wide availability of user-provided content in online social media facilitates the aggregation of people around common interests, worldviews, and narratives. However, the World Wide Web (WWW) also allows for the rapid dissemination of unsubstantiated rumors and conspiracy theories that often elicit rapid, large, but naive social responses such as the recent case of Jade Helm 15––where a simple military exercise turned out to be perceived as the beginning of a new civil war in the United States. In this work, we address the determinants governing misinformation spreading through a thorough quantitative analysis. In particular, we focus on how Facebook users consume information related to two distinct narratives: scientific and conspiracy news. We find that, although consumers of scientific and conspiracy stories present similar consumption patterns with respect to content, cascade dynamics differ. Selective exposure to content is the primary driver of content diffusion and generates the formation of homogeneous clusters, i.e., “echo chambers.” Indeed, homogeneity appears to be the primary driver for the diffusion of contents and each echo chamber has its own cascade dynamics. Finally, we introduce a data-driven percolation model mimicking rumor spreading and we show that homogeneity and polarization are the main determinants for predicting cascades’ size.

HDRF: Stream-Based Partitioning for Power-Law Graphs

Balanced graph partitioning is a fundamental problem that is receiving growing attention with the emergence of distributed graph-computing (DGC) frameworks. In these frameworks, the partitioning strategy plays an important role since it drives the communication cost and the workload balance among computing nodes, thereby affecting system performance. However, existing solutions only partially exploit a key characteristic of natural graphs commonly found in the real-world: their highly skewed power-law degree distributions. In this paper, we propose High-Degree (are) Replicated First (HDRF), a novel streaming vertex-cut graph partitioning algorithm that effectively exploits skewed degree distributions by explicitly taking into account vertex degree in the placement decision. We analytically and experimentally evaluate HDRF on both synthetic and real-world graphs and show that it outperforms all existing algorithms in partitioning quality.

GASGD: stochastic gradient descent for distributed asynchronous matrix completion via graph partitioning.

Matrix completion latent factors models are known to be an effective method to build recommender systems. Currently, stochastic gradient descent (SGD) is considered one of the best latent factor-based algorithm for matrix completion. In this paper we discuss GASGD, a distributed asynchronous variant of SGD for large-scale matrix completion, that (i) leverages data partitioning schemes based on graph partitioning techniques, (ii) exploits specific characteristics of the input data and (iii) introduces an explicit parameter to tune synchronization frequency among the computing nodes. We empirically show how, thanks to these features, GASGD achieves a fast convergence rate incurring in smaller communication cost with respect to current asynchronous distributed SGD implementations.

LCBM: A fast and lightweight collaborative filtering algorithm for binary ratings

Abstract In the last ten years, recommendation systems evolved from novelties to powerful business tools, deeply changing the internet industry. Collaborative Filtering (CF) represents a widely adopted strategy today to build recommendation engines. The most advanced CF techniques (i.e. those based on matrix factorization) provide high quality results, but may incur prohibitive computational costs when applied to very large data sets. In this paper we present Linear Classifier of Beta distributions Means (LCBM), a novel collaborative filtering algorithm for binary ratings that is (i) inherently parallelizable (ii) provides results whose quality is on-par with state-of-the-art solutions (iii) at a fraction of the computational cost. These characteristics allow LCBM to efficiently handle large instances of the collaborative filtering problem on a single machine in short timeframes.

Targeted interest-driven advertising in cities using Twitter

Targeted advertising is a key characteristic of online as well as traditional-media marketing. However it is very limited in outdoor advertising, that is, performing campaigns by means of billboards in public places. The reason is the lack of information about the interests of the particular passersby, except at very imprecise and aggregate demographic or traffic estimates. In this work we propose a methodology for performing targeted outdoor advertising by leveraging the use of social media. In particular, we use the Twitter social network to gather information about users’ degree of interest in given advertising categories and about the common routes that they follow, characterizing in this way each zone in a given city. Then we use our characterization for recommending physical locations for advertising. Given an advertisement category, we estimate the most promising areas to be selected for the placement of an ad that can maximize its targeted effectiveness. We show that our approach is able to select advertising locations better with respect to a baseline reflecting a current ad-placement policy. To the best of our knowledge this is the first work on offline advertising in urban areas making use of (publicly available) data from social networks.

Exploiting user feedback for online filtering in event-based systems

Modern large-scale internet applications represent today a fundamental source of information for millions of users. The larger is the user base, the more difficult it is to control the quality of data that is spread from producers to consumers. This can easily hamper the usability of such systems as the amount of low quality data received by consumers grows uncontrolled. In this paper we propose a novel solution to automatically filter new data injected in event-based systems with the aim of delivering only content consumers are actually interested in. Filtering is executed by profiling producers and consumers, and matching their profiles as new data is produced. Profiles are built by aggregating feedback submitted by consumers on previously received data.

LCBM: Statistics-Based Parallel Collaborative Filtering

In the last ten years, recommendation systems evolved from novelties to powerful business tools, deeply changing the internet industry. Collaborative Filtering (CF) represents today’s a widely adopted strategy to build recommendation engines. The most advanced CF techniques (i.e. those based on matrix factorization) provide high quality results, but may incur prohibitive computational costs when applied to very large data sets. In this paper we present Linear Classifier of Beta distributions Means (LCBM), a novel collaborative filtering algorithm for binary ratings that is (i) inherently parallelizable and (ii) provides results whose quality is on-par with state-of-the-art solutions (iii) at a fraction of the computational cost.

Everyday the same picture: popularity and content diversity

Facebook is flooded by diverse and heterogeneous content, from kittens up to music and news, passing through satirical and funny stories. Each piece of that corpus reflects the heterogeneity of the underlying social background. In the Italian Facebook we have found an interesting case: a page having more than

HSIENA: a hybrid publish/subscribe system

40K

Viral Misinformation: The Role of Homophily and Polarization

followers that every day posts the same picture of a popular Italian singer. In this work, we use such a page as a control to study and model the relationship between content heterogeneity on popularity. In particular, we use that page for a comparative analysis of information consumption patterns with respect to pages posting science and conspiracy news. In total, we analyze about