Telegram emerged as a crucial platform for both parties during the conflict between Russia and Ukraine. Per its minimal policies for content moderation, Pro-Kremlin narratives and potential misinformation were spread on Telegram, while anti-Kremlin narratives with related content were also propagated, such as war footage, troop movements, maps of bomb shelters, and air raid warnings. This paper presents a dataset of posts from both pro-Kremlin and anti-Kremlin Telegram channels, collected over a period spanning a year before and a year after the Russian invasion. The dataset comprises 404 pro-Kremlin channels with 4,109,645 posts and 114 anti-Kremlin channels with 1,117,768 posts. We provide details on the data collection process, processing methods, and dataset characterization. Lastly, we discuss the potential research opportunities this dataset may enable researchers across various disciplines.

The lack of explainability in using relevant clinical knowledge hinders the adoption of artificial intelligence-powered analysis of unstructured clinical dialogue. A wealth of relevant, untapped Mental Health (MH) data is available in online communities, providing the opportunity to address the explainability problem with substantial potential impact as a screening tool for both online and offline applications. Inspired by how clinicians rely on their expertise when interacting with patients, we leverage relevant clinical knowledge to classify and explain depression-related data, reducing manual review time and engendering trust. We developed a method to enhance attention in contemporary transformer models and generate explanations for classifications that are understandable by mental health practitioners (MHPs) by incorporating external clinical knowledge. We propose a domain-general architecture called ProcesS knowledgeinfused cross ATtention (PSAT) that incorporates clinical practice guidelines (CPG) when computing attention. We transform a CPG resource focused on depression, such as the Patient Health Questionnaire (e.g. PHQ-9) and related questions, into a machine-readable ontology using SNOMED-CT. With this resource, PSAT enhances the ability of models like GPT-3.5 to generate application-relevant explanations. Evaluation of four expert-curated datasets related to depression demonstrates PSAT's applicationrelevant explanations. PSAT surpasses the performance of twelve baseline models and can provide explanations where other baselines fall short.

Monitoring public sentiment via social media is potentially helpful during health crises such as the COVID-19 pandemic. However, traditional frequency-based and data-driven neural network-based approaches can miss newly relevant content due to the evolving nature of language in a dynamic environment. Human-curated symbolic knowledge sources, such as lexicons for standard language and slang terms, can potentially elevate social media signals in evolving language. We introduce a neurosymbolic method that integrates neural networks with symbolic knowledge sources, improving the detection and interpretation of mental health-related tweets relevant to COVID-19. Our method was evaluated using a corpus of large datasets (~12 billion tweets, 2.5 million subreddit data, and 700k news articles) and multiple knowledge graphs. This method dynamically adapts to evolving language, outperforming purely data-driven models with an F1 score exceeding 92%. This approach also showed faster adaptation to new data and lower computational demands than fine-tuning pre-trained large language models (LLMs). This study demonstrates the benefit of neurosymbolic methods in interpreting text in a dynamic environment for tasks such as health surveillance.

The digital landscape continually evolves with multimodality, enriching the online experience for users. Creators and marketers aim to weave subtle contextual cues from various modalities into congruent content to engage users with a harmonious message. This interplay of multimodal cues is often a crucial factor in attracting users’ attention. However, this richness of multimodality presents a challenge to computational modeling, as the semantic contextual cues spanning across modalities need to be unified to capture the true holistic meaning of the multimodal content. This contextual meaning is critical in attracting user engagement as it conveys the intended message of the brand or the organization. In this work, we incorporate external commonsense knowledge from knowledge graphs to enhance the representation of multimodal data using compact Visual Language Models (VLMs) and predict the success of multi-modal crowdfunding campaigns. Our results show that external knowledge commonsense bridges the semantic gap between text and image modalities, and the enhanced knowledge-infused representations improve the predictive performance of models for campaign success upon the baselines without knowledge. Our findings highlight the significance of contextual congruence in online multimodal content for engaging and successful crowdfunding campaigns.

The development of highly fluent large language models (LLMs) has prompted increased interest in assessing their reasoning and problem-solving capabilities.We investigate whether several LLMs can solve a classic type of deductive reasoning problem from the cognitive science literature.The tested LLMs have limited abilities to solve these problems in their conventional form.We performed follow up experiments to investigate if changes to the presentation format and content improve model performance.We do find performance differences between conditions; however, they do not improve overall performance.Moreover, we find that performance interacts with presentation format and content in unexpected ways that differ from human performance.Overall, our results suggest that LLMs have unique reasoning biases that are only partially predicted from human reasoning performance and the humangenerated language corpora that informs them.

Common ground processes [26] can improve performance in communication tasks [72, 42, 43, 24], and understanding these processes will likely benefit human--computer dialogue interfaces. However, there are multiple proposed theories with different implications for interface design. Fusaroli and Tylén [40] achieved a direct comparison by designing two models: one based on alignment theory and the other based on complementarity theory that encapsulated interpersonal synergy and audience design. The current research used these models, extending them to differentiate between interpersonal synergy and audience design. Few studies have tested multiple common ground models against tasks representative of envisioned human--computer interaction (HCI) applications. We report on four such tests, which allowed examination of generalizability of findings. Results supported the complementarity models over the alignment model, and were suggestive of the audience design variant of complementarity, providing guidance for HCI design that differs from contemporary approaches.

Dementia is a group of irreversible, chronic, and progressive neurodegenerative disorders resulting in impaired memory, communication, and thought processes. In recent years, clinical research advances in brain aging have focused on the earliest clinically detectable stage of incipient dementia, commonly known as mild cognitive impairment (MCI). Currently, these disorders are diagnosed using a manual analysis of neuropsychological examinations. We measure the feasibility of using the linguistic characteristics of verbal utterances elicited during neuropsychological exams of elderly subjects to distinguish between elderly control groups, people with MCI, people diagnosed with possible Alzheimer's disease (AD) and probable AD. We investigated the performance of both theory-driven psycholinguistic features and data-driven contextual language embeddings in identifying different clinically diagnosed groups. Our experiments show that a combination of contextual and psycholinguistic features extracted by a Support Vector Machine improved distinguishing the verbal utterances of elderly controls, people with MCI, possible AD, and probable AD. This is the first work to identify four clinical diagnosis groups of dementia in a highly imbalanced dataset. Our work shows that machine learning algorithms built on contextual and psycholinguistic features can learn the linguistic biomarkers from verbal utterances and assist clinical diagnosis of different stages and types of dementia, even with limited data.

Imprecise composite location references formed using ad hoc spatial expressions in English text makes the geocoding task challenging for both inference and evaluation. Typically such spatial expressions fill in unestablished areas with new toponyms for finer spatial referents. For example, the spatial extent of the ad hoc spatial expression "north of" or "50 minutes away from" in relation to the toponym "Dayton, OH" refers to an ambiguous, imprecise area, requiring translation from this qualitative representation to a quantitative one with precise semantics using systems such as WGS84. Here we highlight the challenges of geocoding such referents and propose a general formal representation that employs background knowledge, semantic approximations and rules, and fuzzy linguistic variables. We also discuss an appropriate evaluation technique for the task that is based on human contextualized and subjective judgment.

Natural disasters such as floods, forest fires, and hurricanes can cause catastrophic damage to human life and infrastructure. We focus on response to hurricanes caused by both river water flooding and storm surge. Using models for storm surge simulation and flood extent prediction, we generate forecasts about areas likely to be highly affected by the disaster. Further, we overlay the simulation results with information about traffic incidents to correlate traffic incidents with other data modality. We present these results in a modularized, interactive map-based visualization, which can help emergency responders to better plan and coordinate disaster response.

Recent studies show that by combining network topology and node attributes, we can better understand community structures in complex networks. However, existing algorithms do not explore "contextually" similar node attribute values, and therefore may miss communities defined with abstract concepts. We propose a community detection and characterization algorithm that incorporates the contextual information of node attributes described by multiple domain-specific hierarchical concept graphs. The core problem is to find the context that can best summarize the nodes in communities, while also discovering communities aligned with the context summarizing communities. We formulate the two intertwined problems, optimal community-context computation, and community discovery, with a coordinate-ascent based algorithm that iteratively updates the nodes' community label assignment with a community-context and computes the best context summarizing nodes of each community. Our unique contributions include (1) a composite metric on Informativeness and Purity criteria in searching for the best context summarizing nodes of a community; (2) a node similarity measure that incorporates the context-level similarity on multiple node attributes; and (3) an integrated algorithm that drives community structure discovery by appropriately weighing edges. Experimental results on public datasets show nearly 20 percent improvement on F-measure and Jaccard for discovering underlying community structure over the current state-of-the-art of community detection methods. Community structure characterization was also accurate to find appropriate community types for four datasets.

Terror attacks have been linked in part to online extremist content. Online conversations are cloaked in religious ambiguity, with deceptive intentions, often twisted from mainstream meaning to serve a malevolent ideology. Although tens of thousands of Islamist extremism supporters consume such content, they are a small fraction relative to peaceful Muslims. The efforts to contain the ever-evolving extremism on social media platforms have remained inadequate and mostly ineffective. Divergent extremist and mainstream contexts challenge machine interpretation, with a particular threat to the precision of classification algorithms. Radicalization is a subtle long-running persuasive process that occurs over time. Our context-aware computational approach to the analysis of extremist content on Twitter breaks down this persuasion process into building blocks that acknowledge inherent ambiguity and sparsity that likely challenge both manual and automated classification. Based on prior empirical and qualitative research in social sciences, particularly political science, we model this process using a combination of three contextual dimensions -- religion, ideology, and hate -- each elucidating a degree of radicalization and highlighting independent features to render them computationally accessible. We utilize domain-specific knowledge resources for each of these contextual dimensions such as Qur'an for religion, the books of extremist ideologues and preachers for political ideology and a social media hate speech corpus for hate. The significant sensitivity of the Islamist extremist ideology and its local and global security implications require reliable algorithms for modelling such communications on Twitter. Our study makes three contributions to reliable analysis: (i) Development of a computational approach rooted in the contextual dimensions of religion, ideology, and hate, which reflects strategies employed by online Islamist extremist groups, (ii) An in-depth analysis of relevant tweet datasets with respect to these dimensions to exclude likely mislabeled users, and (iii) A framework for understanding online radicalization as a process to assist counter-programming. Given the potentially significant social impact, we evaluate the performance of our algorithms to minimize mislabeling, where our context-aware approach outperforms a competitive baseline by 10.2% in precision, thereby enhancing the potential of such tools for use in human review.

The assessment of trust between users is essential for collaboration. General reputation and ID mechanisms may support users’ trust assessment. However, these mechanisms lack sensitivity to pairwise interactions and specific experience such as betrayal over time. Moreover, they place an interpretation burden that does not scale to dynamic, large-scale systems. While several pairwise trust mechanisms have been proposed, no empirical research examines trust score influence on participant behavior. We study the influence of showing a partner trust score and/or ID on participants’ behavior in a small-group collaborative laboratory experiment based on the trust game. We show that trust score availability has the same effect as an ID to improve cooperation as measured by sending behavior and receiver response. Excellent models based on the trust score predict sender behavior and document participant sensitivity to the provision of partner information. Models based on the trust score for recipient behavior have some predictive ability regarding trustworthiness, but suggest the need for more complex functions relating experience to participant response. We conclude that the parameters of a trust score, including pairwise interactions and betrayal, influence the different roles of participants in the trust game differently, but complement traditional ID and have the advantage of scalability.

Extracting location names from informal and unstructured social media data requires the identification of referent boundaries and partitioning compound names. Variability, particularly systematic variability in location names (Carroll, 1983), challenges the identification task. Some of this variability can be anticipated as operations within a statistical language model, in this case drawn from gazetteers such as OpenStreetMap (OSM), Geonames, and DBpedia. This permits evaluation of an observed n-gram in Twitter targeted text as a legitimate location name variant from the same location-context. Using n-gram statistics and location-related dictionaries, our Location Name Extraction tool (LNEx) handles abbreviations and automatically filters and augments the location names in gazetteers (handling name contractions and auxiliary contents) to help detect the boundaries of multi-word location names and thereby delimit them in texts. We evaluated our approach on 4,500 event-specific tweets from three targeted streams to compare the performance of LNEx against that of ten state-of-the-art taggers that rely on standard semantic, syntactic and/or orthographic features. LNEx improved the average F-Score by 33-179%, outperforming all taggers. Further, LNEx is capable of stream processing.

A crowd sampled from a set of individuals can provide a more accurate prediction in aggregate than most individuals.This effect, referred to as wisdom of crowd, exists when crowd members bring diverse perspectives to decision making. Such diversity leads to uncorrelated prediction errors that cancel out in aggregate. As crowd members' judgments are often the result of solution strategies, diversity in solution strategies can enhance crowd wisdom. One of the most challenging tasks in sampling such a crowd is to determine the individual's solution strategy for a prediction problem. As participating individuals often share their perspectives through social media, we can use such data to identify an individual's solution strategy. In this paper, we propose a crowd selection approach using social media posts (tweets) indicating diverse solution strategies. We use tweet classification to identify participants' prediction strategies and categorize participants based on the binomial test to identify sets of participants that apply a similar strategy. We then form a diverse crowd by sampling participants from different sets. Using the domain of Fantasy Sports, we show that such a diverse crowd can outperform crowd selected at random and 90% of individual participants, and participant categorization schemes using word2vec. Further, we use a knowledge graph to investigate the factors forming such a diverse crowd and how these factors can lead to a better decision. Relative to bottom-up (data-driven) processes the approach presented here provides an explanation of diverse crowd behavior.

"Wisdom of Crowds" (WoC) refers to a form of collective intelligence in which the aggregate judgment of a group of individuals is, in most instances, superior to that of any one group member. For a crowd to be wise, its members must possess diverse knowledge and viewpoints. Such diversity leads to uncorrelated judgment errors that cancel out in aggregate. Yet despite the fact that diversity is known to be an essential ingredient in WoC, little research aims to measure and exploit diversity in human social systems for the purpose of maximizing crowd intelligence. Here we quantify the diversity of a group of individuals through semantic analysis of their social media (Twitter) communications. Focusing on the domain of fantasy sports, we show that virtual crowds of fantasy team owners selected based on the diversity of their tweet content can outperform both non-diverse and randomly sampled crowds. Our results suggest a new approach for intelligent crowd assembly in which measures of diversity extracted from online social media communications can guide the selection of crowd members. These results have implications for numerous domains that utilize aggregated judgments - from consumer reviews, to econometrics, to geopolitical forecasting and intelligence analysis.

Hypertext systems parse documents into content nodes connected by machine supported links or relationships. Many hypertext researchers claim that the node-link relationships of hypertext provide an information organization that models the structure of human knowledge and should therefore facilitate information access (Fiderio 1988). Yet, failures of information access occur when users lack an understanding of the overall scope and organization of a hypertext system (Gay and Mazur 1991). To support this understanding, the present research incorporated expert-based domain semantics in the design of prosthetic devices for hypertext navigation. The task domain was documentation for a word processing system. In the first experiment, the pathfinder algorithm (Schvaneveldt 1990) and cluster analysis were used to identify a set of expertbased semantic relationships between word-processing concepts. The results from these analyses contributed to the design of two prostheses to assist hypertext navigation: A hierarchical index and a local semantic browser. These aids were tested in a second experiment, crossing type of on-line documentation (semantically enhanced hypertext or an alphabetically indexed text) with level of subject expertise (novice or expert). Both performance and strategy measures suggest that the semantic prostheses improved the accessibility of information for novice users without hampering expert performance.

This paper examines physician expertise in a hospital medical intensive care unit and illustrates three functions that expert physicians perform : (1) pursuit of ill-structured goals ; (2) treatment of dynamic disease processes in individual patients ; and (3) detection and ownership of problematic circumstances. These functions of medical expertise are illustrated using examples from videotapes of physician activities collected over five days in an inner city teaching hospital, providing the basis for the following hypotheses. Goal-directed reasoning in this domain extends beyond the technical foundations of medicine to the cultural values that acknowledge its expertise (Agnew et al. 1994). The treatment of dynamic disease processes depends on the adaptation of accepted methods to individual differences among patients, according to known symptoms and diagnoses. Diagnosis is a subgoal to treatment, pursued at multiple levels of abstraction, and only when it is not superseded by more urgent treatment goals. Problem ownership—the acceptance of a problem as medical and suitable for treatment—requires medical knowledge and an awareness of the distribution of technical knowledge and responsibilities throughout a health care organization. These hypotheses provide a functional specification of physician expertise, and the foundation for a test of proficiency claimed by humans or machines.

No specific guidelines exist to assist in designing usable hypertext systems. In this paper, we discuss three experiments to study usability issues in hypertext design. In the first experiment, we investigated usability of four types of hypertext interfaces: graphical with labeled links (GL), graphical with unlabeled links (GU), textual with embedded links (TE), and textual with a separate list of related items/links (TS). The results favored GL interface for novice users. However, most subjects suggested incorporating multiple access pathways to facilitate search. To determine how hypertext designers could establish, a priori, these multiple structures, we extracted organization schemes from domain experts in the second experiment. Distinctly different organization structures emerged from experts with different professional backgrounds. Therefore, we modified the hypertext to incorporate multiple organization structures. In experiment 3, we compared subjects’ performance using multiple and single organization structures. Multiple structures, contrary to previous evidence, enhanced search performance. The benefits of multiple structures, however, diminished over time. These experiments provide empirical evidence in favor of GL interfaces and incorporation of multiple organization structures to improve hypertext usability.