arXiv subjects

Inwon Kang

Publications and source records attributed to Inwon Kang.

10 recordsLinked to original sources

arXiv2026

SemStruct: Contextualizing Semantic Embeddings with Structural Information for Schema Matching

Schema matching is a fundamental step in integrating heterogeneous data sources. While Pre-trained Language Models (PLMs) have revolutionized this task by capturing linguistic semantics, they typically process tabular data as serialized text sequences of standalone column descriptions. This serialization discards critical structural information -- specifically, the row-level co-occurrences, i.e. the relational context -- forcing models to rely solely on column header semantics or standalone distributions. To bridge this gap, we propose SemStruct, a framework that joins the semantic power of frozen PLMs with the structural inductive bias of Graph Neural Networks (GNNs). We model the table as a heterogeneous graph where columns and values are nodes connected by rows, allowing the GNN to propagate disambiguating context across the structure. Unlike other state-of-the-art methods that require proprietary LLM access and fine-tuning of language models, SemStruct keeps the language model frozen and trains only a lightweight structural encoder. Extensive experiments on the Valentine and SOTAB-SM benchmarks demonstrate that SemStruct achieves state-of-the-art performance, outperforming fully fine-tuned baselines on complex, semantically joinable datasets. Furthermore, our ablation studies reveal that row representations serve primarily as topological conduits rather than semantic entities, validating the necessity of explicit structural modeling in schema matching.

cs.LG↗

arXiv2026

Measuring Privacy Risks and Tradeoffs in Financial Synthetic Data Generation

We explore the privacy-utility tradeoff of synthetic data generation schemes on tabular financial datasets, a domain characterized by high regulatory risk and severe class imbalance. We consider representative tabular data generators, including autoencoders, generative adversarial networks, diffusion, and copula synthesizers. To address the challenges of the financial domain, we provide novel privacy-preserving implementations of GAN and autoencoder synthesizers. We evaluate whether and how well the generators simultaneously achieve data quality, downstream utility, and privacy, with comparison across balanced and imbalanced input datasets. Our results offer insight into the distinct challenges of generating synthetic data from datasets that exhibit severe class imbalance and mixed-type attributes.

cs.LG↗

arXiv2026

Language Model Representations for Efficient Few-Shot Tabular Classification

The Web is a rich source of structured data in the form of tables, from product catalogs and knowledge bases to scientific datasets. However, the heterogeneity of the structure and semantics of these tables makes it challenging to build a unified method that can effectively leverage the information they contain. Meanwhile, Large language models (LLMs) are becoming an increasingly integral component of web infrastructure for tasks like semantic search. This raises a crucial question: can we leverage these already-deployed LLMs to classify structured data in web-native tables (e.g., product catalogs, knowledge base exports, scientific data portals), avoiding the need for specialized models or extensive retraining? This work investigates a lightweight paradigm, $\textbf{Ta}$ble $\textbf{R}$epresentation with $\textbf{L}$anguage Model~($\textbf{TaRL}$), for few-shot tabular classification that directly utilizes semantic embeddings of individual table rows. We first show that naive application of these embeddings underperforms compared to specialized tabular models. We then demonstrate that their potentials can be unlocked with two key techniques: removing the common component from all embeddings and calibrating the softmax temperature. We show that a simple meta-learner, trained on handcrafted features, can learn to predict an appropriate temperature. This approach achieves performance comparable to state-of-the-art models in low-data regimes ($k \leq 32$) of semantically-rich tables. Our findings demonstrate the viability of reusing existing LLM infrastructure for efficient semantics-driven pathway to reuse existing LLM infrastructure for Web table understanding.

cs.CL↗

arXiv2025

Terminators: Terms of Service Parsing and Auditing Agents

Terms of Service (ToS) documents are often lengthy and written in complex legal language, making them difficult for users to read and understand. To address this challenge, we propose Terminators, a modular agentic framework that leverages large language models (LLMs) to parse and audit ToS documents. Rather than treating ToS understanding as a black-box summarization problem, Terminators breaks the task down to three interpretable steps: term extraction, verification, and accountability planning. We demonstrate the effectiveness of our method on the OpenAI ToS using GPT-4o, highlighting strategies to minimize hallucinations and maximize auditability. Our results suggest that structured, agent-based LLM workflows can enhance both the usability and enforceability of complex legal documents. By translating opaque terms into actionable, verifiable components, Terminators promotes ethical use of web content by enabling greater transparency, empowering users to understand their digital rights, and supporting automated policy audits for regulatory or civic oversight.

cs.IR↗

arXiv2025

On Learning Representations for Tabular Data Distillation

Dataset distillation generates a small set of information-rich instances from a large dataset, resulting in reduced storage requirements, privacy or copyright risks, and computational costs for downstream modeling, though much of the research has focused on the image data modality. We study tabular data distillation, which brings in novel challenges such as the inherent feature heterogeneity and the common use of non-differentiable learning models (such as decision tree ensembles and nearest-neighbor predictors). To mitigate these challenges, we present $\texttt{TDColER}$, a tabular data distillation framework via column embeddings-based representation learning. To evaluate this framework, we also present a tabular data distillation benchmark, ${{\sf \small TDBench}}$. Based on an elaborate evaluation on ${{\sf \small TDBench}}$, resulting in 226,890 distilled datasets and 548,880 models trained on them, we demonstrate that $\texttt{TDColER}$ is able to boost the distilled data quality of off-the-shelf distillation schemes by 0.5-143% across 7 different tabular learning models.

cs.LG↗

arXiv2024

Enhancing Web Spam Detection through a Blockchain-Enabled Crowdsourcing Mechanism

The proliferation of spam on the Web has necessitated the development of machine learning models to automate their detection. However, the dynamic nature of spam and the sophisticated evasion techniques employed by spammers often lead to low accuracy in these models. Traditional machine-learning approaches struggle to keep pace with spammers' constantly evolving tactics, resulting in a persistent challenge to maintain high detection rates. To address this, we propose blockchain-enabled incentivized crowdsourcing as a novel solution to enhance spam detection systems. We create an incentive mechanism for data collection and labeling by leveraging blockchain's decentralized and transparent framework. Contributors are rewarded for accurate labels and penalized for inaccuracies, ensuring high-quality data. A smart contract governs the submission and evaluation process, with participants staking cryptocurrency as collateral to guarantee integrity. Simulations show that incentivized crowdsourcing improves data quality, leading to more effective machine-learning models for spam detection. This approach offers a scalable and adaptable solution to the challenges of traditional methods.

cs.CR↗

arXiv2024

Using Large Language Models for Generating Smart Contracts for Health Insurance from Textual Policies

We explore using Large Language Models (LLMs) to generate application code that automates health insurance processes from text-based policies. We target blockchain-based smart contracts as they offer immutability, verifiability, scalability, and a trustless setting: any number of parties can use the smart contracts, and they need not have previously established trust relationships with each other. Our methodology generates outputs at increasing levels of technical detail: (1) textual summaries, (2) declarative decision logic, and (3) smart contract code with unit tests. We ascertain LLMs are good at the task (1), and the structured output is useful to validate tasks (2) and (3). Declarative languages (task 2) are often used to formalize healthcare policies, but their execution on blockchain is non-trivial. Hence, task (3) attempts to directly automate the process using smart contracts. To assess the LLM output, we propose completeness, soundness, clarity, syntax, and functioning code as metrics. Our evaluation employs three health insurance policies (scenarios) with increasing difficulty from Medicare's official booklet. Our evaluation uses GPT-3.5 Turbo, GPT-3.5 Turbo 16K, GPT-4, GPT-4 Turbo and CodeLLaMA. Our findings confirm that LLMs perform quite well in generating textual summaries. Although outputs from tasks (2)-(3) are useful starting points, they require human oversight: in multiple cases, even "runnable" code will not yield sound results; the popularity of the target language affects the output quality; and more complex scenarios still seem a bridge too far. Nevertheless, our experiments demonstrate the promise of LLMs for translating textual process descriptions into smart contracts.

cs.CL↗

arXiv2024

Deciphering Crypto Twitter

Cryptocurrency is a fast-moving space, with a continuous influx of new projects every year. However, an increasing number of incidents in the space, such as hacks and security breaches, threaten the growth of the community and the development of technology. This dynamic and often tumultuous landscape is vividly mirrored and shaped by discussions within Crypto Twitter, a key digital arena where investors, enthusiasts, and skeptics converge, revealing real-time sentiments and trends through social media interactions. We present our analysis on a Twitter dataset collected during a formative period of the cryptocurrency landscape. We collected 40 million tweets using cryptocurrency-related keywords and performed a nuanced analysis that involved grouping the tweets by semantic similarity and constructing a tweet and user network. We used sentence-level embeddings and autoencoders to create K-means clusters of tweets and identified six groups of tweets and their topics to examine different cryptocurrency-related interests and the change in sentiment over time. Moreover, we discovered sentiment indicators that point to real-life incidents in the crypto world, such as the FTX incident of November 2022. We also constructed and analyzed different networks of tweets and users in our dataset by considering the reply and quote relationships and analyzed the largest components of each network. Our networks reveal a structure of bot activity in Crypto Twitter and suggest that they can be detected and handled using a network-based approach. Our work sheds light on the potential of social media signals to detect and understand crypto events, benefiting investors, regulators, and curious observers alike, as well as the potential for bot detection in Crypto Twitter using a network-based approach.

cs.CE↗

arXiv2023

LLM-augmented Preference Learning from Natural Language

Finding preferences expressed in natural language is an important but challenging task. State-of-the-art(SotA) methods leverage transformer-based models such as BERT, RoBERTa, etc. and graph neural architectures such as graph attention networks. Since Large Language Models (LLMs) are equipped to deal with larger context lengths and have much larger model sizes than the transformer-based model, we investigate their ability to classify comparative text directly. This work aims to serve as a first step towards using LLMs for the CPC task. We design and conduct a set of experiments that format the classification task into an input prompt for the LLM and a methodology to get a fixed-format response that can be automatically evaluated. Comparing performances with existing methods, we see that pre-trained LLMs are able to outperform the previous SotA models with no fine-tuning involved. Our results show that the LLMs can consistently outperform the SotA when the target text is large -- i.e. composed of multiple sentences --, and are still comparable to the SotA performance in shorter text. We also find that few-shot learning yields better performance than zero-shot learning.

cs.CL↗

arXiv2022

Blockchain Interoperability Landscape

Blockchain has become a popular emergent technology in many industries. It is suitable for a broad range of applications, from its base role as an immutable distributed ledger to the deployment of distributed applications. Many organizations are adopting the technology, but choosing a specific blockchain implementation in an emerging field exposes them to significant technology risk. Selecting the wrong implementation could expose an organization to security vulnerabilities, reduce access to its target audience, or cause issues in the future when switching to a more mature protocol. Blockchain interoperability aims to solve this adaptability problem by increasing the extensibility of blockchain, enabling the addition of new use cases and features without sacrificing the performance of the original blockchain. However, most existing blockchain platforms need to be designed for interoperability, and simple operations like sending assets across platforms create problems. Cryptographic protocols that are secure in isolation may become insecure when several different (individually secure) protocols are composed. Similarly, utilizing trusted custodians may undercut most of the benefits of decentralization offered by blockchain-based systems. Even though there is some research and development in the field of blockchain interoperability, a characterization of the interoperability solutions for various infrastructure options is lacking. This paper presents a methodology for characterizing blockchain interoperability solutions that will help focus on new developments and evaluate existing and future solutions in this space.

cs.CR↗