arXiv subjects

Jisun An

Publications and source records attributed to Jisun An.

At least 37 records · Page 2Linked to original sources

arXiv2025

A semantic embedding space based on large language models for modelling human beliefs

Beliefs form the foundation of human cognition and decision-making, guiding our actions and social connections. A model encapsulating beliefs and their interrelationships is crucial for understanding their influence on our actions. However, research on belief interplay has often been limited to beliefs related to specific issues and relied heavily on surveys. We propose a method to study the nuanced interplay between thousands of beliefs by leveraging an online user debate data and mapping beliefs onto a neural embedding space constructed using a fine-tuned large language model (LLM). This belief space captures the interconnectedness and polarization of diverse beliefs across social issues. Our findings show that positions within this belief space predict new beliefs of individuals and estimate cognitive dissonance based on the distance between existing and new beliefs. This study demonstrates how LLMs, combined with collective online records of human beliefs, can offer insights into the fundamental principles that govern human belief formation.

cs.CL↗

arXiv2024

ToBlend: Token-Level Blending With an Ensemble of LLMs to Attack AI-Generated Text Detection

The robustness of AI-content detection models against sophisticated adversarial strategies, such as paraphrasing or word switching, is a rising concern in natural language generation (NLG) applications. This study proposes ToBlend, a novel token-level ensemble text generation method to challenge the robustness of current AI-content detection approaches by utilizing multiple sets of candidate generative large language models (LLMs). By randomly sampling token(s) from candidate LLMs sets, we find ToBlend significantly drops the performance of most mainstream AI-content detection methods. We evaluate the text quality produced under different ToBlend settings based on annotations from experienced human experts. We proposed a fine-tuned Llama3.1 model to distinguish the ToBlend generated text more accurately. Our findings underscore our proposed text generation approach's great potential in deceiving and improving detection models. Our datasets, codes, and annotations are open-sourced.

cs.CL↗

arXiv2024

Enhancing Stance Classification on Social Media Using Quantified Moral Foundations

This study enhances stance detection on social media by incorporating deeper psychological attributes, specifically individuals' moral foundations. These theoretically-derived dimensions aim to provide a comprehensive profile of an individual's moral concerns which, in recent work, has been linked to behaviour in a range of domains, including society, politics, health, and the environment. In this paper, we investigate how moral foundation dimensions can contribute to predicting an individual's stance on a given target. Specifically we incorporate moral foundation features extracted from text, along with message semantic features, to classify stances at both message- and user-levels using both traditional machine learning models and large language models. Our preliminary results suggest that encoding moral foundations can enhance the performance of stance detection tasks and help illuminate the associations between specific moral foundations and online stances on target topics. The results highlight the importance of considering deeper psychological attributes in stance analysis and underscores the role of moral foundations in guiding online social behavior.

cs.CL↗

arXiv2024

Spatio-Temporal Road Traffic Prediction using Real-time Regional Knowledge

For traffic prediction in transportation services such as car-sharing and ride-hailing, mid-term road traffic prediction (within a few hours) is considered essential. However, the existing road-level traffic prediction has mainly studied how significantly micro traffic events propagate to the adjacent roads in terms of short-term prediction. On the other hand, recent attempts have been made to incorporate regional knowledge such as POIs, road characteristics, and real-time social events to help traffic prediction. However, these studies lack in understandings of different modalities of road-level and region-level spatio-temporal correlations and how to combine such knowledge. This paper proposes a novel method that embeds real-time region-level knowledge using POIs, satellite images, and real-time LTE access traces via a regional spatio-temporal module that consists of dynamic convolution and temporal attention, and conducts bipartite spatial transform attention to convert into road-level knowledge. Then the model ingests this embedded knowledge into a road-level attention-based prediction model. Experimental results on real-world road traffic prediction show that our model outperforms the baselines.

cs.AI↗

arXiv2024

Multiple Areal Feature Aware Transportation Demand Prediction

A reliable short-term transportation demand prediction supports the authorities in improving the capability of systems by optimizing schedules, adjusting fleet sizes, and generating new transit networks. A handful of research efforts incorporate one or a few areal features while learning spatio-temporal correlation, to capture similar demand patterns between similar areas. However, urban characteristics are polymorphic, and they need to be understood by multiple areal features such as land use, sociodemographics, and place-of-interest (POI) distribution. In this paper, we propose a novel spatio-temporal multi-feature-aware graph convolutional recurrent network (ST-MFGCRN) that fuses multiple areal features during spatio-temproal understanding. Inside ST-MFGCRN, we devise sentinel attention to calculate the areal similarity matrix by allowing each area to take partial attention if the feature is not useful. We evaluate the proposed model on two real-world transportation datasets, one with our constructed BusDJ dataset and one with benchmark TaxiBJ. Results show that our model outperforms the state-of-the-art baselines up to 7\% on BusDJ and 8\% on TaxiBJ dataset.

cs.AI↗

arXiv2024

Can we trust the evaluation on ChatGPT?

ChatGPT, the first large language model (LLM) with mass adoption, has demonstrated remarkable performance in numerous natural language tasks. Despite its evident usefulness, evaluating ChatGPT's performance in diverse problem domains remains challenging due to the closed nature of the model and its continuous updates via Reinforcement Learning from Human Feedback (RLHF). We highlight the issue of data contamination in ChatGPT evaluations, with a case study of the task of stance detection. We discuss the challenge of preventing data contamination and ensuring fair model evaluation in the age of closed and continuously trained models.

cs.CL↗

arXiv2024

X-posing Free Speech: Examining the Impact of Moderation Relaxation on Online Social Networks

We investigate the impact of free speech and the relaxation of moderation on online social media platforms using Elon Musk's takeover of Twitter as a case study. By curating a dataset of over 10 million tweets, our study employs a novel framework combining content and network analysis. Our findings reveal a significant increase in the distribution of certain forms of hate content, particularly targeting the LGBTQ+ community and liberals. Network analysis reveals the formation of cohesive hate communities facilitated by influential bridge users, with substantial growth in interactions hinting at increased hate production and diffusion. By tracking the temporal evolution of PageRank, we identify key influencers, primarily self-identified far-right supporters disseminating hate against liberals and woke culture. Ironically, embracing free speech principles appears to have enabled hate speech against the very concept of freedom of expression and free speech itself. Our findings underscore the delicate balance platforms must strike between open expression and robust moderation to curb the proliferation of hate online.

cs.SI↗

arXiv2024

Rematch: Robust and Efficient Matching of Local Knowledge Graphs to Improve Structural and Semantic Similarity

Knowledge graphs play a pivotal role in various applications, such as question-answering and fact-checking. Abstract Meaning Representation (AMR) represents text as knowledge graphs. Evaluating the quality of these graphs involves matching them structurally to each other and semantically to the source text. Existing AMR metrics are inefficient and struggle to capture semantic similarity. We also lack a systematic evaluation benchmark for assessing structural similarity between AMR graphs. To overcome these limitations, we introduce a novel AMR similarity metric, rematch, alongside a new evaluation for structural similarity called RARE. Among state-of-the-art metrics, rematch ranks second in structural similarity; and first in semantic similarity by 1--5 percentage points on the STS-B and SICK-R benchmarks. Rematch is also five times faster than the next most efficient metric.

cs.CL↗

arXiv2024

ChatGPT Rates Natural Language Explanation Quality Like Humans: But on Which Scales?

As AI becomes more integral in our lives, the need for transparency and responsibility grows. While natural language explanations (NLEs) are vital for clarifying the reasoning behind AI decisions, evaluating them through human judgments is complex and resource-intensive due to subjectivity and the need for fine-grained ratings. This study explores the alignment between ChatGPT and human assessments across multiple scales (i.e., binary, ternary, and 7-Likert scale). We sample 300 data instances from three NLE datasets and collect 900 human annotations for both informativeness and clarity scores as the text quality measurement. We further conduct paired comparison experiments under different ranges of subjectivity scores, where the baseline comes from 8,346 human annotations. Our results show that ChatGPT aligns better with humans in more coarse-grained scales. Also, paired comparisons and dynamic prompting (i.e., providing semantically similar examples in the prompt) improve the alignment. This research advances our understanding of large language models' capabilities to assess the text explanation quality in different configurations for responsible AI development.

cs.CL↗

arXiv2024

Benchmarking zero-shot stance detection with FlanT5-XXL: Insights from training data, prompting, and decoding strategies into its near-SoTA performance

We investigate the performance of LLM-based zero-shot stance detection on tweets. Using FlanT5-XXL, an instruction-tuned open-source LLM, with the SemEval 2016 Tasks 6A, 6B, and P-Stance datasets, we study the performance and its variations under different prompts and decoding strategies, as well as the potential biases of the model. We show that the zero-shot approach can match or outperform state-of-the-art benchmarks, including fine-tuned models. We provide various insights into its performance including the sensitivity to instructions and prompts, the decoding strategies, the perplexity of the prompts, and to negations and oppositions present in prompts. Finally, we ensure that the LLM has not been trained on test datasets, and identify a positivity bias which may partially explain the performance differences across decoding strategie

cs.CL↗

arXiv2023

Improving Real Estate Appraisal with POI Integration and Areal Embedding

Despite advancements in real estate appraisal methods, this study primarily focuses on two pivotal challenges. Firstly, we explore the often-underestimated impact of Points of Interest (POI) on property values, emphasizing the necessity for a comprehensive, data-driven approach to feature selection. Secondly, we integrate road-network-based Areal Embedding to enhance spatial understanding for real estate appraisal. We first propose a revised method for POI feature extraction, and discuss the impact of each POI for house price appraisal. Then we present the Areal embedding-enabled Masked Multihead Attention-based Spatial Interpolation for House Price Prediction (AMMASI) model, an improvement upon the existing ASI model, which leverages masked multi-head attention on geographic neighbor houses and similar-featured houses. Our model outperforms current baselines and also offers promising avenues for future optimization in real estate appraisal methodologies.

cs.AI↗

arXiv2023

Enhancing Spatiotemporal Traffic Prediction through Urban Human Activity Analysis

Traffic prediction is one of the key elements to ensure the safety and convenience of citizens. Existing traffic prediction models primarily focus on deep learning architectures to capture spatial and temporal correlation. They often overlook the underlying nature of traffic. Specifically, the sensor networks in most traffic datasets do not accurately represent the actual road network exploited by vehicles, failing to provide insights into the traffic patterns in urban activities. To overcome these limitations, we propose an improved traffic prediction method based on graph convolution deep learning algorithms. We leverage human activity frequency data from National Household Travel Survey to enhance the inference capability of a causal relationship between activity and traffic patterns. Despite making minimal modifications to the conventional graph convolutional recurrent networks and graph convolutional transformer architectures, our approach achieves state-of-the-art performance without introducing excessive computational overhead.

cs.LG↗

arXiv2023

Public Perception of Generative AI on Twitter: An Empirical Study Based on Occupation and Usage

The emergence of generative AI has sparked substantial discussions, with the potential to have profound impacts on society in all aspects. As emerging technologies continue to advance, it is imperative to facilitate their proper integration into society, managing expectations and fear. This paper investigates users' perceptions of generative AI using 3M posts on Twitter from January 2019 to March 2023, especially focusing on their occupation and usage. We find that people across various occupations, not just IT-related ones, show a strong interest in generative AI. The sentiment toward generative AI is generally positive, and remarkably, their sentiments are positively correlated with their exposure to AI. Among occupations, illustrators show exceptionally negative sentiment mainly due to concerns about the unethical usage of artworks in constructing AI. People use ChatGPT in diverse ways, and notably the casual usage in which they "play with" ChatGPT tends to associate with positive sentiments. After the release of ChatGPT, people's interest in AI in general has increased dramatically; however, the topic with the most significant increase and positive sentiment is related to crypto, indicating the hype-worthy characteristics of generative AI. These findings would offer valuable lessons for policymaking on the emergence of new technology and also empirical insights for the considerations of future human-AI symbiosis.

cs.SI↗

arXiv2023

YouNICon: YouTube's CommuNIty of Conspiracy Videos

Conspiracy theories are widely propagated on social media. Among various social media services, YouTube is one of the most influential sources of news and entertainment. This paper seeks to develop a dataset, YOUNICON, to enable researchers to perform conspiracy theory detection as well as classification of videos with conspiracy theories into different topics. YOUNICON is a dataset with a large collection of videos from suspicious channels that were identified to contain conspiracy theories in a previous study (Ledwich and Zaitsev 2020). Overall, YOUNICON will enable researchers to study trends in conspiracy theories and understand how individuals can interact with the conspiracy theory producing community or channel. Our data is available at: https://doi.org/10.5281/zenodo.7466262.

cs.CY↗

arXiv2023

Wearing Masks Implies Refuting Trump?: Towards Target-specific User Stance Prediction across Events in COVID-19 and US Election 2020

People who share similar opinions towards controversial topics could form an echo chamber and may share similar political views toward other topics as well. The existence of such connections, which we call connected behavior, gives researchers a unique opportunity to predict how one would behave for a future event given their past behaviors. In this work, we propose a framework to conduct connected behavior analysis. Neural stance detection models are trained on Twitter data collected on three seemingly independent topics, i.e., wearing a mask, racial equality, and Trump, to detect people's stance, which we consider as their online behavior in each topic-related event. Our results reveal a strong connection between the stances toward the three topical events and demonstrate the power of past behaviors in predicting one's future behavior.

cs.SI↗

arXiv2023

Chain of Explanation: New Prompting Method to Generate Higher Quality Natural Language Explanation for Implicit Hate Speech

Recent studies have exploited advanced generative language models to generate Natural Language Explanations (NLE) for why a certain text could be hateful. We propose the Chain of Explanation (CoE) Prompting method, using the heuristic words and target group, to generate high-quality NLE for implicit hate speech. We improved the BLUE score from 44.0 to 62.3 for NLE generation by providing accurate target information. We then evaluate the quality of generated NLE using various automatic metrics and human annotations of informativeness and clarity scores.

cs.CL↗

arXiv2023

Is ChatGPT better than Human Annotators? Potential and Limitations of ChatGPT in Explaining Implicit Hate Speech

Recent studies have alarmed that many online hate speeches are implicit. With its subtle nature, the explainability of the detection of such hateful speech has been a challenging problem. In this work, we examine whether ChatGPT can be used for providing natural language explanations (NLEs) for implicit hateful speech detection. We design our prompt to elicit concise ChatGPT-generated NLEs and conduct user studies to evaluate their qualities by comparison with human-written NLEs. We discuss the potential and limitations of ChatGPT in the context of implicit hateful speech research.

cs.CL↗

arXiv2023

Political Honeymoon Effect on Social Media: Characterizing Social Media Reaction to the Changes of Prime Minister in Japan

New leaders in democratic countries typically enjoy high approval ratings immediately after taking office. This phenomenon is called the honeymoon effect and is regarded as a significant political phenomenon; however, its mechanism remains underexplored. Therefore, this study examines how social media users respond to changes in political leadership in order to better understand the honeymoon effect in politics. In particular, we constructed a 15-year Twitter dataset on eight change timings of Japanese prime ministers consisting of 6.6M tweets and analyzed them in terms of sentiments, topics, and users. We found that, while not always, social media tend to show a honeymoon effect at the change timings of prime minister. The study also revealed that sentiment about prime ministers differed by topic, indicating that public expectations vary from one prime minister to another. Furthermore, the user base was largely replaced before and after the change in the prime minister, and their sentiment was also significantly different. The implications of this study would be beneficial for administrative management.

cs.SI↗