arXiv subjects

Shanika Karunasekera

Publications and source records attributed to Shanika Karunasekera.

At least 19 recordsLinked to original sources

arXiv2026

Revisiting the Objective of Echo Chamber Detection

In this paper, we study the detection of an echo chamber in a social network, i.e., the identification of a set of nodes that agree on a topic, while disagreeing with the rest of nodes. We argue that this problem is different from other social network analysis problems such as community detection, and from other graph problems such as maximum graph cut and maximum clique. To the best of our knowledge, we are the first to formalize the objective function of echo chamber detection, by using the theory of Fourier transforms of set functions (Stobbe and Krause, 2012). We propose scalable semidefinite relaxation, solved via an interior point method and sparse linear algebra. Experimentally, our algorithm recovers the ground truth echo chamber better than competing methods on small synthetic experiments. Our algorithm produces echo chambers with better network properties than competing methods on large real-world datasets. To independently validate our proposed objective function, we show that our algorithm finds echo chambers with more agreements with suspended users than competing methods on a small real-world dataset.

cs.LG↗

arXiv2026

UrbanVerse: Learning Urban Region Representation Across Cities and Tasks

Recent advances in urban region representation learning have enabled a wide range of applications in urban analytics, yet existing methods remain limited in their capabilities to generalize across cities and analytic tasks. We aim to generalize urban representation learning beyond city- and task-specific settings, towards a foundation-style model for urban analytics. To this end, we propose UrbanVerse, a model for cross-city urban representation learning and cross-task urban analytics. For cross-city generalization, UrbanVerse focuses on features local to the target regions and structural features of the nearby regions rather than the entire city. We model regions as nodes on a graph, which enables a random walk-based procedure to form "sequences of regions" that reflect both local and neighborhood structural features for urban region representation learning. For cross-task generalization, we propose a cross-task learning module named HCondDiffCT. This module integrates region-conditioned prior knowledge and task-conditioned semantics into the diffusion process to jointly model multiple downstream urban prediction tasks. HCondDiffCT is generic. It can also be integrated with existing urban representation learning models to enhance their downstream task effectiveness. Experiments on real-world datasets show that UrbanVerse consistently outperforms state-of-the-art methods across six tasks under cross-city settings, achieving up to 35.89% improvements in prediction accuracy.

cs.LG↗

arXiv2025

Dynamic Configuration of On-Street Parking Spaces using Multi Agent Reinforcement Learning

With increased travelling needs more than ever, traffic congestion has become a major concern in most urban areas. Allocating spaces for on-street parking, further hinders traffic flow, by limiting the effective road width available for driving. With the advancement of vehicle-to-infrastructure connectivity technologies, we explore how the impact of on-street parking on traffic congestion could be minimized, by dynamically configuring on-street parking spaces. Towards that end, we formulate dynamic on-street parking space configuration as an optimization problem, and we follow a data driven approach, considering the nature of our problem. Our proposed solution comprises a two-layer multi agent reinforcement learning based framework, which is inherently scalable to large road networks. The lane level agents are responsible for deciding the optimal parking space configuration for each lane, and we introduce a novel Deep Q-learning architecture which effectively utilizes long short term memory networks and graph attention networks to capture the spatio-temporal correlations evident in the given problem. The block level agents control the actions of the lane level agents and maintain a sufficient level of parking around the block. We conduct a set of comprehensive experiments using SUMO, on both synthetic data as well as real-world data from the city of Melbourne. Our experiments show that the proposed framework could reduce the average travel time loss of vehicles significantly, reaching upto 47%, with a negligible increase in the walking distance for parking.

cs.LG↗

arXiv2025

FlexiReg: Flexible Urban Region Representation Learning

The increasing availability of urban data offers new opportunities for learning region representations, which can be used as input to machine learning models for downstream tasks such as check-in or crime prediction. While existing solutions have produced promising results, an issue is their fixed formation of regions and fixed input region features, which may not suit the needs of different downstream tasks. To address this limitation, we propose a model named FlexiReg for urban region representation learning that is flexible with both the formation of urban regions and the input region features. FlexiReg is based on a spatial grid partitioning over the spatial area of interest. It learns representations for the grid cells, leveraging publicly accessible data, including POI, land use, satellite imagery, and street view imagery. We propose adaptive aggregation to fuse the cell representations and prompt learning techniques to tailor the representations towards different tasks, addressing the needs of varying formations of urban regions and downstream tasks. Extensive experiments on five real-world datasets demonstrate that FlexiReg outperforms state-of-the-art models by up to 202% in term of the accuracy of four diverse downstream tasks using the produced urban region representations.

cs.LG↗

arXiv2025

Social Correction on Social Media: A Quantitative Analysis of Comment Behaviour and Reliability

Corrections given by ordinary social media users, also referred to as Social Correction have emerged as a viable intervention against misinformation as per the recent literature. However, little is known about how often users give disputing or endorsing comments and how reliable those comments are. An online experiment was conducted to investigate how users' credibility evaluations of social media posts and their confidence in those evaluations combined with online reputational concerns affect their commenting behaviour. The study found that participants exhibited a more conservative approach when giving disputing comments compared to endorsing ones. Nevertheless, participants were more discerning in their disputing comments than endorsing ones. These findings contribute to a better understanding of social correction on social media and highlight the factors influencing comment behaviour and reliability.

cs.SI↗

arXiv2025

Langformers: Unified NLP Pipelines for Language Models

Transformer-based language models have revolutionized the field of natural language processing (NLP). However, using these models often involves navigating multiple frameworks and tools, as well as writing repetitive boilerplate code. This complexity can discourage non-programmers and beginners, and even slow down prototyping for experienced developers. To address these challenges, we introduce Langformers, an open-source Python library designed to streamline NLP pipelines through a unified, factory-based interface for large language model (LLM) and masked language model (MLM) tasks. Langformers integrates conversational AI, MLM pretraining, text classification, sentence embedding/reranking, data labelling, semantic search, and knowledge distillation into a cohesive API, supporting popular platforms such as Hugging Face and Ollama. Key innovations include: (1) task-specific factories that abstract training, inference, and deployment complexities; (2) built-in memory and streaming for conversational agents; and (3) lightweight, modular design that prioritizes ease of use. Documentation: https://langformers.com

cs.CL↗

arXiv2025

Large Language Model Driven Agents for Simulating Echo Chamber Formation

The rise of echo chambers on social media platforms has heightened concerns about polarization and the reinforcement of existing beliefs. Traditional approaches for simulating echo chamber formation have often relied on predefined rules and numerical simulations, which, while insightful, may lack the nuance needed to capture complex, real-world interactions. In this paper, we present a novel framework that leverages large language models (LLMs) as generative agents to simulate echo chamber dynamics within social networks. The novelty of our approach is that it incorporates both opinion updates and network rewiring behaviors driven by LLMs, allowing for a context-aware and semantically rich simulation of social interactions. Additionally, we utilize real-world Twitter (now X) data to benchmark the LLM-based simulation against actual social media behaviors, providing insights into the accuracy and realism of the generated opinion trends. Our results demonstrate the efficacy of LLMs in modeling echo chamber formation, capturing both structural and semantic dimensions of opinion clustering. %This work contributes to a deeper understanding of social influence dynamics and offers a new tool for studying polarization in online communities.

cs.SI↗

arXiv2025

"Actionable Help" in Crises: A Novel Dataset and Resource-Efficient Models for Identifying Request and Offer Social Media Posts

During crises, social media serves as a crucial coordination tool, but the vast influx of posts--from "actionable" requests and offers to generic content like emotional support, behavioural guidance, or outdated information--complicates effective classification. Although generative LLMs (Large Language Models) can address this issue with few-shot classification, their high computational demands limit real-time crisis response. While fine-tuning encoder-only models (e.g., BERT) is a popular choice, these models still exhibit higher inference times in resource-constrained environments. Moreover, although distilled variants (e.g., DistilBERT) exist, they are not tailored for the crisis domain. To address these challenges, we make two key contributions. First, we present CrisisHelpOffer, a novel dataset of 101k tweets collaboratively labelled by generative LLMs and validated by humans, specifically designed to distinguish actionable content from noise. Second, we introduce the first crisis-specific mini models optimized for deployment in resource-constrained settings. Across 13 crisis classification tasks, our mini models surpass BERT (also outperform or match the performance of RoBERTa, MPNet, and BERTweet), offering higher accuracy with significantly smaller sizes and faster speeds. The Medium model is 47% smaller with 3.8% higher accuracy at 3.5x speed, the Small model is 68% smaller with a 1.8% accuracy gain at 7.7x speed, and the Tiny model, 83% smaller, matches BERT's accuracy at 18.6x speed. All models outperform existing distilled variants, setting new benchmarks. Finally, as a case study, we analyze social media posts from a global crisis to explore help-seeking and assistance-offering behaviours in selected developing and developed countries.

cs.CL↗

arXiv2024

CReMa: Crisis Response through Computational Identification and Matching of Cross-Lingual Requests and Offers Shared on Social Media

During times of crisis, social media platforms play a crucial role in facilitating communication and coordinating resources. In the midst of chaos and uncertainty, communities often rely on these platforms to share urgent pleas for help, extend support, and organize relief efforts. However, the overwhelming volume of conversations during such periods can escalate to unprecedented levels, necessitating the automated identification and matching of requests and offers to streamline relief operations. Additionally, there is a notable absence of studies conducted in multi-lingual settings, despite the fact that any geographical area can have a diverse linguistic population. Therefore, we propose CReMa (Crisis Response Matcher), a systematic approach that integrates textual, temporal, and spatial features to address the challenges of effectively identifying and matching requests and offers on social media platforms during emergencies. Our approach utilizes a crisis-specific pre-trained model and a multi-lingual embedding space. We emulate human decision-making to compute temporal and spatial features and non-linearly weigh the textual features. The results from our experiments are promising, outperforming strong baselines. Additionally, we introduce a novel multi-lingual dataset simulating help-seeking and offering assistance on social media in 16 languages and conduct comprehensive cross-lingual experiments. Furthermore, we analyze a million-scale geotagged global dataset to understand patterns in seeking help and offering assistance on social media. Overall, these contributions advance the field of crisis informatics and provide benchmarks for future research in the area.

cs.CL↗

arXiv2024

Using Causality to Infer Coordinated Attacks in Social Media

The rise of social media has been accompanied by a dark side with the ease of creating fake accounts and disseminating misinformation through coordinated attacks. Existing methods to identify such attacks often rely on thematic similarities or network-based approaches, overlooking the intricate causal relationships that underlie coordinated actions. This work introduces a novel approach for detecting coordinated attacks using Convergent Cross Mapping (CCM), a technique that infers causality from temporal relationships between user activity. We build on the theoretical framework of CCM by incorporating topic modelling as a basis for further optimizing its performance. We apply CCM to real-world data from the infamous IRA attack on US elections, achieving F1 scores up to 75.3% in identifying coordinated accounts. Furthermore, we analyse the output of our model to identify the most influential users in a community. We apply our model to a case study involving COVID-19 anti-vax related discussions on Twitter. Our results demonstrate the effectiveness of our model in uncovering causal structures of coordinated behaviour, offering a promising avenue for mitigating the threat of malicious campaigns on social media platforms.

cs.SI↗

arXiv2024

Identifying Coordinated Activities on Online Social Networks Using Contrast Pattern Mining

The proliferation of misinformation and disinformation on social media networks has become increasingly concerning. With a significant portion of the population using social media on a regular basis, there are growing efforts by malicious organizations to manipulate public opinion through coordinated campaigns. Current methods for identifying coordinated user accounts typically rely on either similarities in user behaviour, latent coordination in activity traces, or classification techniques. In our study, we propose a framework based on the hypothesis that coordinated users will demonstrate abnormal growth in their behavioural patterns over time relative to the wider population. Specifically, we utilize the EPClose algorithm to extract contrasting patterns of user behaviour during a time window of malicious activity, which we then compare to a historical time window. We evaluated the effectiveness of our approach using real-world data, and our results show a minimum increase of 10% in the F1 score compared to existing approaches.

cs.SI↗

arXiv2024

To be a pro-vax or not, the COVID-19 vaccine conundrum on Twitter

The most surprising observation reported by the study in (arXiv:2208.13523), involving stance detection of COVID-19 vaccine related tweets during the first year of pandemic, is the presence of a significant number of users (~2 million) who posted tweets with both anti-vax and pro-vax stances. This is a sizable cohort even when the stance detection noise is considered. In this paper, we tried to get deeper understanding of this 'dual-stance' group. Out of this group, 60% of users have more pro-vax tweets than anti-vax tweets and 17% have the same number of tweets in both classes. The rest have more anti-vax tweets, and they were highly active in expressing concerns about mandate and safety of a fast-tracked vaccine, while also tweeted some updates about vaccine development. The leaning pro-vax group have opposite composition: more vaccine updates and some posts about concerns. It is important to note that vaccine concerns were not always genuine and had a large dose of misinformation. 43% of the balanced group have only tweeted one tweet of each type during our study period and are the less active participants in the vaccine discourse. Our temporal study also shows that the change-of-stance behaviour became really significant once the trial results of COVID-19 vaccine were announced to the public, and it appears as the change of stance towards pro-vax is a reaction to people changing their opinion towards anti-vax. Our study finished at Mar 23, 2021 when the conundrum was still going strong. The dilemma might be a reflection of the uncertain and stressful times, but it also highlights the importance of building public trust to combat prevalent misinformation.

cs.SI↗

arXiv2024

Urban Region Representation Learning with Attentive Fusion

An increasing number of related urban data sources have brought forth novel opportunities for learning urban region representations, i.e., embeddings. The embeddings describe latent features of urban regions and enable discovering similar regions for urban planning applications. Existing methods learn an embedding for a region using every different type of region feature data, and subsequently fuse all learned embeddings of a region to generate a unified region embedding. However, these studies often overlook the significance of the fusion process. The typical fusion methods rely on simple aggregation, such as summation and concatenation, thereby disregarding correlations within the fused region embeddings. To address this limitation, we propose a novel model named HAFusion. Our model is powered by a dual-feature attentive fusion module named DAFusion, which fuses embeddings from different region features to learn higher-order correlations between the regions as well as between the different types of region features. DAFusion is generic - it can be integrated into existing models to enhance their fusion process. Further, motivated by the effective fusion capability of an attentive module, we propose a hybrid attentive feature learning module named HALearning to enhance the embedding learning from each individual type of region features. Extensive experiments on three real-world datasets demonstrate that our model HAFusion outperforms state-of-the-art methods across three different prediction tasks. Using our learned region embedding leads to consistent and up to 31% improvements in the prediction accuracy.

cs.LG↗

arXiv2024

A Systematic Literature Review on Reasons and Approaches for Accurate Effort Estimations in Agile

Background: Accurate effort estimation is crucial for planning in Agile iterative development. Agile estimation generally relies on consensus-based methods like planning poker, which require less time and information than other formal methods (e.g., COSMIC) but are prone to inaccuracies. Understanding the common reasons for inaccurate estimations and how proposed approaches can assist practitioners is essential. However, prior systematic literature reviews (SLR) only focus on the estimation practices (e.g., [26, 127]) and the effort estimation approaches (e.g., [6]). Aim: We aim to identify themes of reasons for inaccurate estimations and classify approaches to improve effort estimation. Method: We conducted an SLR and identified the key themes and a taxonomy. Results: The reasons for inaccurate estimation are related to information quality, team, estimation practice, project management, and business influences. The effort estimation approaches were the most investigated in the literature, while only a few aim to support the effort estimation process. Yet, few automated approaches are at risk of data leakage and indirect validation scenarios. Recommendations: Practitioners should enhance the quality of information for effort estimation, potentially by adopting an automated approach. Future research should aim to improve the information quality, while avoiding data leakage and indirect validation scenarios.

cs.SE↗

arXiv2024

CrisisTransformers: Pre-trained language models and sentence encoders for crisis-related social media texts

Social media platforms play an essential role in crisis communication, but analyzing crisis-related social media texts is challenging due to their informal nature. Transformer-based pre-trained models like BERT and RoBERTa have shown success in various NLP tasks, but they are not tailored for crisis-related texts. Furthermore, general-purpose sentence encoders are used to generate sentence embeddings, regardless of the textual complexities in crisis-related texts. Advances in applications like text classification, semantic search, and clustering contribute to the effective processing of crisis-related texts, which is essential for emergency responders to gain a comprehensive view of a crisis event, whether historical or real-time. To address these gaps in crisis informatics literature, this study introduces CrisisTransformers, an ensemble of pre-trained language models and sentence encoders trained on an extensive corpus of over 15 billion word tokens from tweets associated with more than 30 crisis events, including disease outbreaks, natural disasters, conflicts, and other critical incidents. We evaluate existing models and CrisisTransformers on 18 crisis-specific public datasets. Our pre-trained models outperform strong baselines across all datasets in classification tasks, and our best-performing sentence encoder improves the state-of-the-art by 17.43% in sentence encoding tasks. Additionally, we investigate the impact of model initialization on convergence and evaluate the significance of domain-specific models in generating semantically meaningful sentence embeddings. The models are publicly available at: https://huggingface.co/crisistransformers

cs.CL↗

arXiv2024

Semantically Enriched Cross-Lingual Sentence Embeddings for Crisis-related Social Media Texts

Tasks such as semantic search and clustering on crisis-related social media texts enhance our comprehension of crisis discourse, aiding decision-making and targeted interventions. Pre-trained language models have advanced performance in crisis informatics, but their contextual embeddings lack semantic meaningfulness. Although the CrisisTransformers family includes a sentence encoder to address the semanticity issue, it remains monolingual, processing only English texts. Furthermore, employing separate models for different languages leads to embeddings in distinct vector spaces, introducing challenges when comparing semantic similarities between multi-lingual texts. Therefore, we propose multi-lingual sentence encoders (CT-XLMR-SE and CT-mBERT-SE) that embed crisis-related social media texts for over 50 languages, such that texts with similar meanings are in close proximity within the same vector space, irrespective of language diversity. Results in sentence encoding and sentence matching tasks are promising, suggesting these models could serve as robust baselines when embedding multi-lingual crisis-related social media texts. The models are publicly available at: https://huggingface.co/crisistransformers.

cs.CL↗

arXiv2023

A Deep Reinforcement Learning based Algorithm for Time and Cost Optimized Scaling of Serverless Applications

Serverless computing has gained a strong traction in the cloud computing community in recent years. Among the many benefits of this novel computing model, the rapid auto-scaling capability of user applications takes prominence. However, the offer of adhoc scaling of user deployments at function level introduces many complications to serverless systems. The added delay and failures in function request executions caused by the time consumed for dynamically creating new resources to suit function workloads, known as the cold-start delay, is one such very prevalent shortcoming. Maintaining idle resource pools to alleviate this issue often results in wasted resources from the cloud provider perspective. Existing solutions to address this limitation mostly focus on predicting and understanding function load levels in order to proactively create required resources. Although these solutions improve function performance, the lack of understanding on the overall system characteristics in making these scaling decisions often leads to the sub-optimal usage of system resources. Further, the multi-tenant nature of serverless systems requires a scalable solution adaptable for multiple co-existing applications, a limitation seen in most current solutions. In this paper, we introduce a novel multi-agent Deep Reinforcement Learning based intelligent solution for both horizontal and vertical scaling of function resources, based on a comprehensive understanding on both function and system requirements. Our solution elevates function performance reducing cold starts, while also offering the flexibility for optimizing resource maintenance cost to the service providers. Experiments conducted considering varying workload scenarios show improvements of up to 23% and 34% in terms of application latency and request failures, while also saving up to 45% in infrastructure cost for the service providers.

cs.DC↗

arXiv2023

Unsupervised Domain-agnostic Fake News Detection using Multi-modal Weak Signals

The emergence of social media as one of the main platforms for people to access news has enabled the wide dissemination of fake news. This has motivated numerous studies on automating fake news detection. Although there have been limited attempts at unsupervised fake news detection, their performance suffers due to not exploiting the knowledge from various modalities related to news records and due to the presence of various latent biases in the existing news datasets. To address these limitations, this work proposes an effective framework for unsupervised fake news detection, which first embeds the knowledge available in four modalities in news records and then proposes a novel noise-robust self-supervised learning technique to identify the veracity of news records from the multi-modal embeddings. Also, we propose a novel technique to construct news datasets minimizing the latent biases in existing news datasets. Following the proposed approach for dataset construction, we produce a Large-scale Unlabelled News Dataset consisting 419,351 news articles related to COVID-19, acronymed as LUND-COVID. We trained the proposed unsupervised framework using LUND-COVID to exploit the potential of large datasets, and evaluate it using a set of existing labelled datasets. Our results show that the proposed unsupervised framework largely outperforms existing unsupervised baselines for different tasks such as multi-modal fake news detection, fake news early detection and few-shot fake news detection, while yielding notable improvements for unseen domains during training.

cs.LG↗