The COVID-19 pandemic has fueled the spread of misinformation on social media and the Web as a whole. The phenomenon dubbed `infodemic' has taken the challenges of information veracity and trust to new heights by massively introducing seemingly scientific and technical elements into misleading content. Despite the existing body of work on modeling and predicting misinformation, the coverage of very complex scientific topics with inherent uncertainty and an evolving set of findings, such as COVID-19, provides many new challenges that are not easily solved by existing tools. To address these issues, we introduce SciLander, a method for learning representations of news sources reporting on science-based topics. We extract four heterogeneous indicators for the sources; two generic indicators that capture (1) the copying of news stories between sources, and (2) the use of the same terms to mean different things (semantic shift), and two scientific indicators that capture (1) the usage of jargon and (2) the stance towards specific citations. We use these indicators as signals of source agreement, sampling pairs of positive (similar) and negative (dissimilar) samples, and combine them in a unified framework to train unsupervised news source embeddings with a triplet margin loss objective. We evaluate our method on a novel COVID-19 dataset containing nearly 1M news articles from 500 sources spanning a period of 18 months since the beginning of the pandemic in 2020. Our results show that the features learned by our model outperform state-of-the-art baseline methods on the task of news veracity classification. Furthermore, a clustering analysis suggests that the learned representations encode information about the reliability, political leaning, and partisanship bias of these sources.

Integrity 2023 is the fourth edition of the successful Workshop on Integrity in Social Networks and Media, held in conjunction with the ACM Conference on Web Search and Data Mining (WSDM) in the past three years. The goal of the workshop is to bring together researchers and practitioners to discuss content and interaction integrity challenges in social networks and social media platforms. The event consists of a combination of invited talks by reputed members of the Integrity community from both academia and industry and peer-reviewed contributed talks and posters solicited via an open call-for-papers.

Lexical Semantic Change (LSC) detection, also known as Semantic Shift, is the process of identifying and characterizing variations in language usage across different scenarios such as time and domain. It allows us to track the evolution of word senses, as well as to understand the difference between the language used in distinct communities. LSC detection is often done by applying a distance measure over vectors of two aligned word embedding matrices. In this demonstration, we present SenSE, an interactive semantic shift exploration toolkit that provides visualization and explanation of lexical semantic change for an input pair of text sources. Our system focuses on showing how the different alignment strategies may affect the output of an LSC model as well as on explaining semantic change based on the neighbors of a chosen target word, while also extracting examples of sentences where these semantic deviations appear. The system runs as a web application (available at http://sense.mgruppi.me), allowing the audience to interact by configuring the alignment strategies while visualizing the results in a web browser.

The use of language is subject to variation over time as well as across social groups and knowledge domains, leading to differences even in the monolingual scenario. Such variation in word usage is often called lexical semantic change (LSC). The goal of LSC is to characterize and quantify language variations with respect to word meaning, to measure how distinct two language sources are (that is, people or language models). Because there is hardly any data available for such a task, most solutions involve unsupervised methods to align two embeddings and predict semantic change with respect to a distance measure. To that end, we propose a self-supervised approach to model lexical semantic change based on the perturbation of word vectors in the input corpora. We show that our method can be used for the detection of semantic change with any alignment method. Furthermore, it can be used to choose the landmark words to use in alignment and can lead to substantial improvements over the existing techniques for alignment. We illustrate the utility of our techniques using experimental results on three different datasets, involving words with the same or different meanings. Our methods not only provide significant improvements but also can lead to novel findings for the LSC problem.

In this study, we examine the impact of time on state-of-the-art news veracity classifiers. We show that, as time progresses, classification performance for both unreliable and hyper-partisan news classification slowly degrade. While this degradation does happen, it happens slower than expected, illustrating that hand-crafted, content-based features, such as style of writing, are fairly robust to changes in the news cycle. We show that this small degradation can be mitigated using online learning. Last, we examine the impact of adversarial content manipulation by malicious news producers. Specifically, we test three types of attack based on changes in the input space and data availability. We show that static models are susceptible to content manipulation attacks, but online models can recover from such attacks.

We extend the Top-Trading-Cycles (TTC) mechanism to select strict core allocations for housing markets with multiple types of items, where each agent may be endowed and allocated with multiple items of each type. In doing so, we advance the state of the art in mechanism design for housing markets along two dimensions: First, our setting is more general than multi-type housing markets (Moulin 1995; Sikdar, Adali, and Xia 2017) and the setting of Fujita et al. (2015). Further, we introduce housing markets with acceptable bundles (HMABs) as a more general setting where each agent may have arbitrary sets of acceptable bundles. Second, our extension of TTC is strict core selecting under the weaker restriction on preferences of CMI-trees, which we introduce as a new domain restriction on preferences that generalizes commonly-studied languages in previous works.

We study how emotions influence norm outcomes in decision-making contexts. Following the literature, we provide baseline Dynamic Bayesian models to capture an agent's two perspectives on a directed norm. Unlike the literature, these models are holistic in that they incorporate not only norm outcomes and emotions but also trust and goals. We obtain data from an empirical study involving game play with respect to the above variables. We provide a step-wise process to discover two new Dynamic Bayesian models based on maximizing log-likelihood scores with respect to the data. We compare the new models with the baseline models to discover new insights into the relevant relationships. Our empirically supported models are thus holistic and characterize how emotions influence norm outcomes better than previous approaches.

The proliferation of information disseminated by public/social media has made decision-making highly challenging due to the wide availability of noisy, uncertain, or unverified information. Although the issue of uncertainty in information has been studied for several decades, little work has investigated how noisy (or uncertain) or valuable (or credible) information can be formulated into people's opinions, modeling uncertainty both in the quantity and quality of evidence leading to a specific opinion. In this work, we model and analyze an opinion and information model by using Subjective Logic where the initial set of evidence is mixed with different types of evidence (i.e., pro vs. con or noisy vs. valuable) which is incorporated into the opinions of original propagators, who propagate information over a network. With the help of an extensive simulation study, we examine how the different ratios of information types or agents' prior belief or topic competence affect the overall information diffusion. Based on our findings, agents' high uncertainty is not necessarily always bad in making a right decision as long as they are competent enough not to be at least biased towards false information (e.g., neutral between two extremes).

We study multi-type housing markets, where there are p ≥ 2 types of items, each agent is initially endowed one item of each type, and the goal is to design mechanisms without monetary transfer to (re)allocate items to the agents based on their preferences over bundles of items, such that each agent gets one item of each type. In sharp contrast to classical housing markets, previous studies in multi-type housing markets have been hindered by the lack of natural solution concepts, because the strict core might be empty. We break the barrier in the literature by leveraging AI techniques and making natural assumptions on agents’ preferences. We show that when agents’ preferences are lexicographic, even with different importance orders, the classical top-trading-cycles mechanism can be extended while preserving most of its nice properties. We also investigate computational complexity of checking whether an allocation is in the strict core and checking whether the strict core is empty. Our results convey an encouragingly positive message: it is possible to design good mechanisms for multi-type housing markets under natural assumptions on preferences.

To stop or mitigate the dissemination of false information in social networks, many studies have investigated the minimum number of seeding nodes required to significantly reduce the impact of false information. Although a person's confidence level, such as perceived certainty, and/or prior belief towards a given proposition can significantly affect their decision of whether to believe in true or false information, these topics have not been studied to date. In this work, we propose an opinion model based on Subjective Logic (SL), defining an opinion in belief, disbelief, and uncertainty, to study how to eradicate or mitigate the impact of false information by propagating true information to counter it. In the current form of an opinion in SL, when two agents interact with each other and update their opinions based on a consensus operator offered by SL, uncertainty continuously reduces whenever any new information, even conflicting evidence, is received. However, in reality, if a body of evidence is conflicting, with equal amounts supporting opposite positions, people often tend to be confused, leading to higher level of uncertainty. We enhance SL to deal with conflicting information associated with uncertainty. We map agents' opinion composition into each state in the SIR (Susceptible-Infected- Recovered) model to estimate the proportion of recovered agents who believe in true information. Our results show that agents' prior belief unfavoring false information can help guide their decisions towards a belief in true information even under high uncertainty. Further, having more true informers in a network can significantly increase the number of agents who believe in true information and the effect is more pronounced than having more frequent propagation of true information by fewer true informers.

Increasingly people form opinions based on information they consume on online social media. As a result, it is crucial to understand what type of content attracts people's attention on social media and drive discussions. In this paper we focus on online discussions. Can we predict which comments and what content gets the highest attention in an online discussion? How does this content differ from community to community? To accomplish this, we undertake a unique study of Reddit involving a large sample comments from 11 popular subreddits with different properties. We introduce a large number of sentiment, relevance, content analysis features including some novel features customized to reddit. Through a comparative analysis of the chosen subreddits, we show that our models are correctly able to retrieve top replies under a post with great precision. In addition, we explain our findings with a detailed analysis of what distinguishes high scoring posts in different communities that differ along the dimensions of the specificity of topic and style, audience and level of moderation.

In social networks, individuals and systems work side by side. While individuals make decisions to filter or forward information, systems also prioritize and sort information to manage and assist individual information processing. It has long been argued that system level manipulations can reduce access of individuals to novel information. In this paper, we study how sorting of messages in one's inbox can help or hinder access of diverse information in the network through simulation of cognitively bounded actors. We show that first-in-first-out (FIFO) method of message sorting is ideal in bursty information arrival rates and in networks with lower diameter. Last-in-first-out (LIFO) method of message sorting is ideal for streaming information arrival, but leads to information overload in bursty scenarios by creating too many redundant copies of some of the information in the network. In short, the ideal message sorting method that enhances access to diverse information depends on the network type and information access patterns.

Modeling trust in very large social networks is a hard problem due to the highly noisy nature of these networks that span trust relationships from many different contexts, based on judgments of reliability, dependability, and competence. Furthermore, relationships in these networks vary in their level of strength. In this article, we introduce a novel extension of structural balance theory as a foundational theory of trust and distrust in networks. Our theory preserves the distinctions between trust and distrust as suggested in the literature, but also incorporates the notion of relationship strength that can be expressed as either discrete categorical values, as pairwise comparisons, or as metric distances. Our model is novel, has sound social and psychological basis, and captures the classical balance theory as a special case. We then propose a convergence model, describing how an imbalanced network evolves towards new balance, and formulate the convergence problem of a social network as a Metric Multidimensional Scaling (MDS) optimization problem. Finally, we show how the convergence model can be used to predict edge signs in social networks and justify our theory through extensive experiments on real datasets.

We examine whether the prominence of individuals in different social networks is determined by their position in their local network or by how the community to which they belong relates to other communities. To this end, we introduce two new measures of centrality, both based on communities in the network: local and community centrality. Community centrality is a novel concept that we introduce to describe how central one's community is within the whole network. We introduce an algorithm to estimate the distance between communities and use it to find the centrality of communities. Using data from several social networks, we show that community centrality is able to capture the importance of communities in the whole network. We then conduct a detailed study of different social networks and determine how various global measures of prominence relate to structural centrality measures. Our measures deconstruct global centrality along local and community dimensions. In some cases, prominence is determined almost exclusively by local information, while in others a mix of local and community centrality matters. Our methodology is a step toward understanding of the processes that contribute to an actor's prominence in a network.

Modeling trust in very large social networks is a hard problem due to the highly noisy nature of these networks that span trust relationships from many different contexts, based on judgments of reliability, dependability and competence and the relationships vary in their level of strength. In this paper, we introduce a new extended balance theory as a foundational theory of trust in networks. Our theory preserves the distinctions between trust and distrust as suggested in the literature, but also incorporates the notion of relationship strength which can be expressed as either discrete categorical values, as pairwise comparisons or as metric distances. Our model is novel, has sound social and psychological basis, and captures the classical balance theory as a special case. We then propose a convergence model, describing how an imbalanced network evolves towards new balance and formulate the convergence problem of a social network as a Metric Multidimensional Scaling (MDS) optimization problem. Finally, we show how the convergence model can be used to predict edge signs in social networks, and justify our theory through experiments on real datasets.

We present a new algorithm called iHypR for computing prominence of actors in social networks of collaborations. Our algorithm builds on the assumption that prominent actors collaborate on prominent objects, and prominent objects are naturally grouped into prominent clusters or groups (hyperedges in a graph). iHypR makes use of the relationships between actors, objects, and hyperedges to compute a global prominence score for the actors in the network. We do not assume the hyperedges are given in advance. Hyperedges computed by our method can perform as well or even better than “true” hyperedges. Our algorithm is customized for networks of collaborations, but it is generally applicable without further tuning. We show, through extensive experimentation with three real-life data sets and multiple external measures of prominence, that our algorithm outperforms existing well-known algorithms. Our work is the first to offer such an extensive evaluation. We show that unlike most existing algorithms, the performance is robust across multiple measures of performance. Further, we give a detailed study of the sensitivity of our algorithm to different data sets and the design choices within the algorithm that a user may wish to change. Our article illustrates the various trade-offs that must be considered in computing prominence in collaborative social networks.

In this paper, we examine to which degree behavioral measures can be used to predict personality. Personality is one factor that dictates people's propensity to trust and their relationships with others. In previous work, we have shown that personality can be predicted relatively accurately by analyzing social media profiles. We demonstrated this using public data from facebook profiles and text from Twitter streams. As social situations are crucial in the formation of one's personality, one's social behavior could be a strong indicator of her personality. Given most users of social media sites typically have a large number of friends and followers, considering only these aspects may not provide an accurate picture of personality. To overcome this problem, we develop a set of measures based on one's behavior towards her friends and followers. We introduce a number of measures that are based on the intensity and number of social interactions one has with friends along a number of dimensions such as reciprocity and priority. We analyze these features along with a set of features based on the textual analysis of the messages sent by the users. We show that behavioral features are very useful in determining personality and perform as well as textual features.

In recent years, new studies concentrating on analyzing user personality and finding credible content in social media have become quite popular. Most such work augments features from textual content with features representing the user's social ties and the tie strength. Social ties are crucial in understanding the network the people are a part of. However, textual content is extremely useful in understanding topics discussed and the personality of the individual. We bring a new dimension to this type of analysis with methods to compute the type of ties individuals have and the strength of the ties in each dimension. We present a new genre of behavioral features that are able to capture the "function" of a specific relationship without the help of textual features. Our novel features are based on the statistical properties of communication patterns between individuals such as reciprocity, assortativity, attention and latency. We introduce a new methodology for determining how such features can be compared to textual features, and show, using Twitter data, that our features can be used to capture contextual information present in textual features very accurately. Conversely, we also demonstrate how textual features can be used to determine social attributes related to an individual.

Trust is an important yet complex and little understood aspect of the dyadic relationship between two entities. Trust plays an important role in the formation of coalitions in social networks and in determining how high value of information flows through the network. We present algorithmically quantifiable measures of trust based on communication behavior. We propose that trust results in likely communication behaviors which are statistically different from random communications; detecting these trust-like behaviors allows us to develop a quantitative measure of who trusts whom in the network. We develop algorithms to efficiently compute such behavioral trust and validate these measures on the Twitter network.

With the emergence of Web 2.0 applications, where information is not only shared across the internet, but also syndicated, evaluated, selected, recombined, edited, etc., quality emergence by collaborative effort from many users becomes crucial. However, users may have low expertise, subjective views, or competitive goals. Therefore, we need to identify cooperative users with strong expertise and high objectivity.

Over the last few years, there has been a tremendous increase in the number of interactive multimedia presentations prepared by different individuals and organizations. In this paper, we present an algebra for querying multimedia presentation databases. In contrast to the relational algebra, an algebra for interactive multimedia presentations must operate on trees whose branches reflect different possible playouts of a family of presentations. The query language supports selection type operations for locating objects and presentation paths that are of interest to the user, join type operations for combining presentations from multiple databases into a single presentation, and finally set theoretic operations for comparing different databases. The algebra operations can be used to locate presentations with specific properties and also for creating new presentations by borrowing different components from existing ones. We prove a host of equivalence results for queries in this algebra which may be used to build query optimizers for interactive presentation databases.

The aim of information integration is to build sophisticated information systems by making use of the available information sources to the fullest extent and by pushing costly operations to the sources as much as possible. This is especially true when translating queries across multiple multimedia information sources that support advanced and/or similarity based queries. We propose a flexible architecture that allows users to specify a wide range of structured queries using generic and simple query constructs and connectives through a uniform query interface. These queries are translated into resource specific queries by processing rules that specify the properties of different query interfaces as well as the user preferences for evaluating queries. We show how different and multiple notions of query relaxation can be captured in this framework naturally.

The need to automatically extract and classify the contents of multimedia data archives such as images, video, and text documents has led to significant work on similarity based retrieval of data. To date, most work in this area has focused on the creation of index structures for similarity based retrieval. There is very little work on developing formalisms for querying multimedia databases that support similarity based computations and optimizing such queries, even though it is well known that feature extraction and identification algorithms in media data are very expensive. We introduce a similarity algebra that brings together relational operators and results of multiple similarity implementations in a uniform language. The algebra can be used to specify complex queries that combine different interpretations of similarity values and multiple algorithms for computing these values. We prove equivalence and containment relationships between similarity algebra expressions and develop query rewriting methods based on these results. We then provide a generic cost model for evaluating cost of query plans in the similarity algebra and query optimization methods based on this model. We supplement the paper with experimental results that illustrate the use of the algebra and the effectiveness of query optimization methods using the Integrated Search Engine (I.SEE) as the testbed.

Query processing and optimization in mediator systems that access distributed non-proprietary sources pose many novel problems. Cost-based query optimization is hard because the mediator does not have access to source statistics information and furthermore it may not be easy to model the source's performance. At the same time, querying remote sources may be very expensive because of high connection overhead, long computation time, financial charges, and temporary unavailability. We propose a cost-based optimization technique that caches statistics of actual calls to the sources and consequently estimates the cost of the possible execution plans based on the statistics cache. We investigate issues pertaining to the design of the statistics cache and experimentally analyze various tradeoffs. We also present a query result caching mechanism that allows us to effectively use results of prior queries when the source is not readily available. We employ the novel invariants mechanism, which shows how semantic information about data sources may be used to discover cached query results of interest.

Integrating knowledge from multiple sources is an important aspect of automated reasoning systems. Wiederhold and his colleagues (1993) have proposed the concept of a mediator-a device that will express how such an integration is to be achieved. In (1994) Subrahmanian et al. presented a uniform declarative and operational framework for mediators for amalgamating multiple knowledge bases and data structures (e.g. relational, object-oriented, spatial, and temporal structures) when these knowledge bases (possibly) contain inconsistencies, uncertainties, and nonmonotonic modes of negation. We specify the programming environment for this framework and show that it can be used to extract and integrate information obtained from different sources of data and resolve conflicts. We also show that it can be extended easily to integrate new knowledge bases.>

Integrating knowledge from multiple sources is an important aspect of automated reasoning systems. In [23], we presented a uniform declarative and operational framework, based on annotated logics, for amalgamating multiple knowledge bases and data structures (e.g. relational, object-oriented, spatial, and temporal structures) when these knowledge bases (possibly) contain inconsistencies, uncertainties and non-monotonic modes of negation. We showed that annotated logics may be used, with some modifications, to mediate between different knowledge bases. The multiple knowledge bases are amalgamated by embedding the individual knowledge bases into a lattice. In this paper, we describe how, given a network of sites where the different databases reside, it is possible to define a distributed semantics for amalgamated knowledge bases. More importantly, we study how the mediator may be distributed across multiple sites so that when certain conditions are satisfied, network failures do not affect the end results of queries that a user may pose. We specify different ways of distributing the mediator to protect against different types of network link failures and develop alternative soundness and completeness results.