arXiv subjects

Chuan-Ju Wang

Publications and source records attributed to Chuan-Ju Wang.

At least 19 recordsLinked to original sources

arXiv2026

Beyond Uniform Token Training: A Multi-Target Framework for Learning Token-Weighted Objectives in Generative Recommenders

Recent generative recommendation models recast next-item prediction as the generation of a semantic identifier sequence. While this formulation enables autoregressive models to produce item IDs directly, the commonly used token-level likelihood objective does not distinguish between tokens that play different roles in item identification. This limitation is especially pronounced for semantic-ID representations, where prefix tokens often determine coarse item groups and later tokens provide finer-grained disambiguation. To better align training with the structure of semantic IDs, we study token-level learning signals from two complementary perspectives. First, we introduce a prefix-aware weighting scheme, Front-Greater Weighting, which emphasizes tokens according to their contribution to reducing semantic ambiguity among candidate items. Second, frequency weighting increases the learning emphasis on infrequent tokens, addressing the long-tailed distributions and popularity bias commonly observed in recommendation data. We further introduce a multi-target optimization framework with curriculum learning, which integrates the two token-weighted objectives with the standard likelihood and enables stable optimization with adaptive emphasis across training stages. Experiments on multiple benchmark datasets demonstrate that the proposed approach consistently improves generative recommendation performance over strong baselines and prior token-weighting methods. Additional analyses show that the method is robust across different semantic-ID constructions and backbone scales, and that it improves recommendation quality for both popular and long-tail items. Code is available at github repository.

cs.IR↗

arXiv2026

LODESTAR: Robust Entropy-Based Answer Selection in Retrieval-Augmented Generation for Question Answering -- Directing Frozen-LLM Entropy with a Reinforcement-Learned Prompt Polarizer under Misleading Passages

Predictive-distribution entropy is a strong answer-selection rule in retrieval-augmented generation (RAG) for question answering: across five QA benchmarks, selecting the answer a frozen respondent LLM produces with the lowest answer-token entropy lifts mean $F_1$ from 0.4769 to 0.5148 over the retriever's top-ranked passage, without gold answers. Yet this rule, which prior entropy-based selectors adopt, fails: a misleading passage makes the respondent confidently wrong, driving entropy down where the uncertainty signal looks most trustworthy. The failure comes from the passage the respondent reads, and the context it is read in is an input we can intervene on. We introduce LODESTAR: to our knowledge the first method to score a text intervention by the uncertainty it induces in a third-party frozen respondent, compared within one question. LODESTAR uses reinforcement learning (GRPO) to train, once and offline, a polarizer -- a short fixed natural-language string inserted into the respondent's prompt and never into its weights, directing entropy so that entropy-based answer selection stays robust to misleading passages; training labels are built from gold answers and two LLM judges, and inference reads neither. With every competing selector under the same frozen respondent and candidate pools on 5,008 questions, LODESTAR attains the highest mean $F_1$ of any inference-ready selector (0.5339), the highest macro exact match (0.4136), and the highest GPT-4o judge score of the frozen-respondent configurations judged (0.6435); its three-seed mean wins all 70 $F_1$ cells against fourteen published configurations and is paired-significant on $F_1$ against every one. The gain holds in-domain on NQ-Open and out-of-domain over SQuAD, TriviaQA, EntityQuestions and WebQuestions. Ablating the polarizer shows it is what makes the respondent read a misleading passage less often (26.0% vs 30.3%).

cs.CL↗

arXiv2025

Financial Risk Relation Identification through Dual-view Adaptation

A multitude of interconnected risk events -- ranging from regulatory changes to geopolitical tensions -- can trigger ripple effects across firms. Identifying inter-firm risk relations is thus crucial for applications like portfolio management and investment strategy. Traditionally, such assessments rely on expert judgment and manual analysis, which are, however, subjective, labor-intensive, and difficult to scale. To address this, we propose a systematic method for extracting inter-firm risk relations using Form 10-K filings -- authoritative, standardized financial documents -- as our data source. Leveraging recent advances in natural language processing, our approach captures implicit and abstract risk connections through unsupervised fine-tuning based on chronological and lexical patterns in the filings. This enables the development of a domain-specific financial encoder with a deeper contextual understanding and introduces a quantitative risk relation score for transparency, interpretable analysis. Extensive experiments demonstrate that our method outperforms strong baselines across multiple evaluation settings. Our codes are available at https://github.com/cnclabs/codes.fin.relation.

cs.CL↗

arXiv2025

FinNuE: Exposing the Risks of Using BERTScore for Numerical Semantic Evaluation in Finance

BERTScore has become a widely adopted metric for evaluating semantic similarity between natural language sentences. However, we identify a critical limitation: BERTScore exhibits low sensitivity to numerical variation, a significant weakness in finance where numerical precision directly affects meaning (e.g., distinguishing a 2% gain from a 20% loss). We introduce FinNuE, a diagnostic dataset constructed with controlled numerical perturbations across earnings calls, regulatory filings, social media, and news articles. Using FinNuE, demonstrate that BERTScore fails to distinguish semantically critical numerical differences, often assigning high similarity scores to financially divergent text pairs. Our findings reveal fundamental limitations of embedding-based metrics for finance and motivate numerically-aware evaluation frameworks for financial NLP.

cs.CL↗

arXiv2025

CFDA & CLIP at TREC iKAT 2025: Enhancing Personalized Conversational Search via Query Reformulation and Rank Fusion

The 2025 TREC Interactive Knowledge Assistance Track (iKAT) featured both interactive and offline submission tasks. The former requires systems to operate under real-time constraints, making robustness and efficiency as important as accuracy, while the latter enables controlled evaluation of passage ranking and response generation with pre-defined datasets. To address this, we explored query rewriting and retrieval fusion as core strategies. We built our pipelines around Best-of-$N$ selection and Reciprocal Rank Fusion (RRF) strategies to handle different submission tasks. Results show that reranking and fusion improve robustness while revealing trade-offs between effectiveness and efficiency across both tasks.

cs.IR↗

arXiv2025

Test-Time Scaling Strategies for Generative Retrieval in Multimodal Conversational Recommendations

The rapid evolution of e-commerce has exposed the limitations of traditional product retrieval systems in managing complex, multi-turn user interactions. Recent advances in multimodal generative retrieval -- particularly those leveraging multimodal large language models (MLLMs) as retrievers -- have shown promise. However, most existing methods are tailored to single-turn scenarios and struggle to model the evolving intent and iterative nature of multi-turn dialogues when applied naively. Concurrently, test-time scaling has emerged as a powerful paradigm for improving large language model (LLM) performance through iterative inference-time refinement. Yet, its effectiveness typically relies on two conditions: (1) a well-defined problem space (e.g., mathematical reasoning), and (2) the model's ability to self-correct -- conditions that are rarely met in conversational product search. In this setting, user queries are often ambiguous and evolving, and MLLMs alone have difficulty grounding responses in a fixed product corpus. Motivated by these challenges, we propose a novel framework that introduces test-time scaling into conversational multimodal product retrieval. Our approach builds on a generative retriever, further augmented with a test-time reranking (TTR) mechanism that improves retrieval accuracy and better aligns results with evolving user intent throughout the dialogue. Experiments across multiple benchmarks show consistent improvements, with average gains of 14.5 points in MRR and 10.6 points in nDCG@1.

cs.IR↗

arXiv2025

Decision-Oriented Text Evaluation

Natural language generation (NLG) is increasingly deployed in high-stakes domains, yet common intrinsic evaluation methods, such as n-gram overlap or sentence plausibility, weakly correlate with actual decision-making efficacy. We propose a decision-oriented framework for evaluating generated text by directly measuring its influence on human and large language model (LLM) decision outcomes. Using market digest texts--including objective morning summaries and subjective closing-bell analyses--as test cases, we assess decision quality based on the financial performance of trades executed by human investors and autonomous LLM agents informed exclusively by these texts. Our findings reveal that neither humans nor LLM agents consistently surpass random performance when relying solely on summaries. However, richer analytical commentaries enable collaborative human-LLM teams to outperform individual human or agent baselines significantly. Our approach underscores the importance of evaluating generated text by its ability to facilitate synergistic decision-making between humans and LLMs, highlighting critical limitations of traditional intrinsic metrics.

cs.CL↗

arXiv2023

Improving Conversational Passage Re-ranking with View Ensemble

This paper presents ConvRerank, a conversational passage re-ranker that employs a newly developed pseudo-labeling approach. Our proposed view-ensemble method enhances the quality of pseudo-labeled data, thus improving the fine-tuning of ConvRerank. Our experimental evaluation on benchmark datasets shows that combining ConvRerank with a conversational dense retriever in a cascaded manner achieves a good balance between effectiveness and efficiency. Compared to baseline methods, our cascaded pipeline demonstrates lower latency and higher top-ranking effectiveness. Furthermore, the in-depth analysis confirms the potential of our approach to improving the effectiveness of conversational search.

cs.IR↗

arXiv2022

Text-to-Text Multi-view Learning for Passage Re-ranking

Recently, much progress in natural language processing has been driven by deep contextualized representations pretrained on large corpora. Typically, the fine-tuning on these pretrained models for a specific downstream task is based on single-view learning, which is however inadequate as a sentence can be interpreted differently from different perspectives. Therefore, in this work, we propose a text-to-text multi-view learning framework by incorporating an additional view -- the text generation view -- into a typical single-view passage ranking model. Empirically, the proposed approach is of help to the ranking performance compared to its single-view counterpart. Ablation studies are also reported in the paper.

cs.IR↗

arXiv2021

Rényi Divergence in General Hidden Markov Models

In this paper, we examine the existence of the Rényi divergence between two time invariant general hidden Markov models with arbitrary positive initial distributions. By making use of a Markov chain representation of the probability distribution for the general hidden Markov model and eigenvalue for the associated Markovian operator, we obtain, under some regularity conditions, convergence of the Rényi divergence. By using this device, we also characterize the Rényi divergence, and obtain the Kullback-Leibler divergence as α \rightarrow 1 of the Rényi divergence. Several examples, including the classical finite state hidden Markov models, Markov switching models, and recurrent neural networks, are given for illustration. Moreover, we develop a non-Monte Carlo method that computes the Rényi divergence of two-state Markov switching models via the underlying invariant probability measure, which is characterized by the Fredholm integral equation.

cs.IT↗

arXiv2021

Multi-Stage Conversational Passage Retrieval: An Approach to Fusing Term Importance Estimation and Neural Query Rewriting

Conversational search plays a vital role in conversational information seeking. As queries in information seeking dialogues are ambiguous for traditional ad-hoc information retrieval (IR) systems due to the coreference and omission resolution problems inherent in natural language dialogue, resolving these ambiguities is crucial. In this paper, we tackle conversational passage retrieval (ConvPR), an important component of conversational search, by addressing query ambiguities with query reformulation integrated into a multi-stage ad-hoc IR system. Specifically, we propose two conversational query reformulation (CQR) methods: (1) term importance estimation and (2) neural query rewriting. For the former, we expand conversational queries using important terms extracted from the conversational context with frequency-based signals. For the latter, we reformulate conversational queries into natural, standalone, human-understandable queries with a pretrained sequence-tosequence model. Detailed analyses of the two CQR methods are provided quantitatively and qualitatively, explaining their advantages, disadvantages, and distinct behaviors. Moreover, to leverage the strengths of both CQR methods, we propose combining their output with reciprocal rank fusion, yielding state-of-the-art retrieval effectiveness, 30% improvement in terms of NDCG@3 compared to the best submission of TREC CAsT 2019.

cs.CL↗

arXiv2020

Personalized TV Recommendation: Fusing User Behavior and Preferences

In this paper, we propose a two-stage ranking approach for recommending linear TV programs. The proposed approach first leverages user viewing patterns regarding time and TV channels to identify potential candidates for recommendation and then further leverages user preferences to rank these candidates given textual information about programs. To evaluate the method, we conduct empirical studies on a real-world TV dataset, the results of which demonstrate the superior performance of our model in terms of both recommendation accuracy and time efficiency.

cs.IR↗

arXiv2020

Skewness Ranking Optimization for Personalized Recommendation

In this paper, we propose a novel optimization criterion that leverages features of the skew normal distribution to better model the problem of personalized recommendation. Specifically, the developed criterion borrows the concept and the flexibility of the skew normal distribution, based on which three hyperparameters are attached to the optimization criterion. Furthermore, from a theoretical point of view, we not only establish the relation between the maximization of the proposed criterion and the shape parameter in the skew normal distribution, but also provide the analogies and asymptotic analysis of the proposed criterion to maximization of the area under the ROC curve. Experimental results conducted on a range of large-scale real-world datasets show that our model significantly outperforms the state of the art and yields consistently best performance on all tested datasets.

cs.IR↗

arXiv2020

Conversational Question Reformulation via Sequence-to-Sequence Architectures and Pretrained Language Models

This paper presents an empirical study of conversational question reformulation (CQR) with sequence-to-sequence architectures and pretrained language models (PLMs). We leverage PLMs to address the strong token-to-token independence assumption made in the common objective, maximum likelihood estimation, for the CQR task. In CQR benchmarks of task-oriented dialogue systems, we evaluate fine-tuned PLMs on the recently-introduced CANARD dataset as an in-domain task and validate the models using data from the TREC 2019 CAsT Track as an out-domain task. Examining a variety of architectures with different numbers of parameters, we demonstrate that the recent text-to-text transfer transformer (T5) achieves the best results both on CANARD and CAsT with fewer parameters, compared to similar transformer architectures.

cs.CL↗

arXiv2020

TTTTTackling WinoGrande Schemas

We applied the T5 sequence-to-sequence model to tackle the AI2 WinoGrande Challenge by decomposing each example into two input text strings, each containing a hypothesis, and using the probabilities assigned to the "entailment" token as a score of the hypothesis. Our first (and only) submission to the official leaderboard yielded 0.7673 AUC on March 13, 2020, which is the best known result at this time and beats the previous state of the art by over five points.

cs.CL↗

arXiv2019

Efficient Exponential Tilting for Portfolio Credit Risk

This paper considers the problem of measuring the credit risk in portfolios of loans, bonds, and other instruments subject to possible default under multi-factor models. Due to the amount of the portfolio, the heterogeneous effect of obligors, and the phenomena that default events are rare and mutually dependent, it is difficult to calculate portfolio credit risk either by means of direct analysis or crude Monte Carlo under such models. To capture the extreme dependence among obligors, we provide an efficient simulation method for multi-factor models with a normal mixture copula that allows the multivariate defaults to have an asymmetric distribution, while most of the literature focuses on simulating one-dimensional cases. To this end, we first propose a general account of an importance sampling algorithm based on an unconventional exponential embedding, which is related to the classical sufficient statistic. Note that this innovative tilting device is more suitable for the multivariate normal mixture model than traditional one-parameter tilting methods and is of independent interest. Next, by utilizing a fast computational method for how the rare event occurs and the proposed importance sampling method, we provide an efficient simulation algorithm to estimate the probability that the portfolio incurs large losses under the normal mixture copula. Here the proposed simulation device is based on importance sampling for a joint probability other than the conditional probability used in previous studies. Theoretical investigations and simulation studies, which include an empirical example, are given to illustrate the method.

q-fin.CP↗

arXiv2019

Collaborative Similarity Embedding for Recommender Systems

We present collaborative similarity embedding (CSE), a unified framework that exploits comprehensive collaborative relations available in a user-item bipartite graph for representation learning and recommendation. In the proposed framework, we differentiate two types of proximity relations: direct proximity and k-th order neighborhood proximity. While learning from the former exploits direct user-item associations observable from the graph, learning from the latter makes use of implicit associations such as user-user similarities and item-item similarities, which can provide valuable information especially when the graph is sparse. Moreover, for improving scalability and flexibility, we propose a sampling technique that is specifically designed to capture the two types of proximity relations. Extensive experiments on eight benchmark datasets show that CSE yields significantly better performance than state-of-the-art recommendation methods.

cs.IR↗

arXiv2018

Representation Learning for Image-based Music Recommendation

Image perception is one of the most direct ways to provide contextual information about a user concerning his/her surrounding environment; hence images are a suitable proxy for contextual recommendation. We propose a novel representation learning framework for image-based music recommendation that bridges the heterogeneity gap between music and image data; the proposed method is a key component for various contextual recommendation tasks. Preliminary experiments show that for an image-to-song retrieval task, the proposed method retrieves relevant or conceptually similar songs for input images.

cs.MM↗