Cross-Domain Sequential Recommendation (CDSR) seeks to improve user preference modeling by transferring knowledge from multiple domains. Despite the progress made in CDSR, most existing methods rely on overlapping users or items to establish cross-domain correlations-a requirement that rarely holds in real-world settings. The advent of large language models (LLM) and model-merging techniques appears to overcome this limitation by unifying multi-domain data without explicit overlaps. Yet, our empirical study shows that naively training an LLM on combined domains—or simply merging several domain-specific LLMs—often degrades performance relative to a model trained solely on the target domain.

Augmenting large language models (LLMs) with tools significantly enhances their problem-solving potential across multifaceted tasks. However, current tools automatically created by LLMs often serve as a mere summary of specific problems or solutions, which face two main issues: 1) Low reusability: The tools are overly problem-specific and struggle to handle new problems. 2) Limited diversity: The toolsets are too narrow, limiting their application to address a broader range of different problems. In this paper, we propose the Knowledge-grounded Tool Creation with Evolution (KTCE) framework, which aims to craft reusable and comprehensive toolsets for LLMs in a two-stage process. In the first stage (Knowledge-based Tool Creation), we conceptualize tools as a form of executable domain knowledge and propose a problem-knowledge-tool paradigm. Specifically, we leverage LLMs to abstract "knowledge" from "problems" and create a three-layer knowledge tree of topics, concepts, and key points. This hierarchical structure serves as a foundation for inducing atomic "tools" from "knowledge", grounding them in fundamental concepts and enhancing their usability. In the second stage (Tool Evolutionary Search), we evolve the toolsets through several actions including tool selection, mutation, and crossover. This stage mimics the biological evolution process, aiding toolsets in discovering new tools or updating existing ones, thereby increasing the diversity of the toolset. Experiments on challenging mathematical/tabular/scientific reasoning tasks demonstrate that our approach achieves substantial accuracy improvements ranging from 6.23% to 18.49% on average. Moreover, in-depth analyses reveal the superior characteristics of our toolkit, including high reusability, high diversity, and high generalizability on cross-data/LLM performance with low complexity.

Cognitive diagnosis, which assesses the learners' competence from learners' interaction logs, plays a vital role in education. It provides a crucial reference for gauging learners' proficiency levels and tailoring future learning activities accordingly. Researchers have proposed numerous cognitive diagnosis models to address this task. Despite their success, these models continue to face the ill-posed problem because of the information loss caused by under-expressive interaction function and incomplete observations. In this paper, we address these challenges by proposing a novel cognitive diagnosis model, DMC-CDM, based on the theoretical premise that cognitive states can be captured with minimal information loss by maximizing the mutual information between observed and potential observations. Specifically, DMC-CDM incorporates a semantic extractor to provide a comprehensive semantic understanding of learners' interaction logs, thereby enhancing current collaborative-based cognitive state representations. It then consolidates multi-perspective observations to capture precise cognitive states by maximizing mutual information between these observations. We conducted extensive experiments on three datasets, and the experimental results demonstrate that our proposed model is both effective and beneficial for downstream applications in education.

Cheng Cheng, Zhenya Huang, GuanHao Zhao, Yuxiang Guo, Xin Lin, Jinze Wu, Xin Li, Shijin Wang. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Large language models (LLMs) have shown promise in multivariate time series classification (MTSC). To effectively adapt LLMs for MTSC, it is crucial to generate comprehensive and informative data representations. Most methods utilizing LLMs encode numerical time series into the model's latent space, aiming to align with the semantic space of LLMs for more effective learning. Despite effectiveness, we highlight three limitations that these methods overlook: (1) they struggle to incorporate temporal and channel-specific information, both of which are essential components of multivariate time series; (2) aligning the learned representation space with the semantic space of the LLMs proves to be a significant challenge; (3) they often require task-specific retraining, preventing training-free inference despite the generalization capabilities of LLMs. To bridge these gaps, we propose TableTime, which reformulates MTSC as a table understanding task. Specifically, TableTime introduces the following strategies: (1) utilizing tabular form to unify the format of time series, facilitating the transition from the model-centric approach to the data-centric approach; (2) representing time series in text format to facilitate seamless alignment with the semantic space of LLMs; (3) designing a knowledge-task dual-driven reasoning framework, TableTime, integrating contextual information and expert-level reasoning guidance to enhance LLMs' reasoning capabilities and enable training-free classification. Extensive experiments conducted on 10 publicly available benchmark datasets from the UEA archive validate the substantial potential of TableTime to be a new paradigm for MTSC. The code is publicly available. https://github.com/realwangjiahao/TableTime.

Pre-trained foundation models have recently made significant progress in table-related tasks such as table understanding and reasoning. However, recognizing the structure and content of unstructured tables using Vision Large Language Models (VLLMs) remains under-explored. To bridge this gap, we propose a benchmark based on a hierarchical design philosophy to evaluate the recognition capabilities of VLLMs in training-free scenarios. Through in-depth evaluations, we find that low-quality image input is a significant bottleneck in the recognition process. Drawing inspiration from this, we propose the Neighbor-Guided Toolchain Reasoner (NGTR) framework, which is characterized by integrating diverse lightweight tools for visual operations aimed at mitigating issues with low-quality images. Specifically, we transfer a tool selection experience from a similar neighbor to the input and design a reflection module to supervise the tool invocation process. Extensive experiments on public datasets demonstrate that our approach significantly enhances the recognition capabilities of the vanilla VLLMs. We believe that the benchmark and framework could provide an alternative solution to table recognition.

The 2024 Nobel Prizes in Chemistry and Physics have once again drawn global attention to AI for Science. The rise of foundation models has further accelerated AI for Science across multiple disciplines. Scientific research is both a touchstone for advancing the intelligence of these models, and the models themselves are accelerators that empower scientific research. As a high-tech enterprise deeply engaged in artificial intelligence, iFLYTEK has in recent years made AI for Science (AI4S) a strategic priority and undertaken a series of initiatives. In this talk, Dr. Xin Li will provide a comprehensive overview of the iFLYTEK Spark Large Language Model and highlight its recent advances. He will outline two principal pathways through which AI accelerates scientific research-deep neural networks and LLM-based approaches-and present iFLYTEK's work along both lines. In addition, he will discuss key challenges and the outlook for the future development of AI for Science. Attendees will gain insights into how AI empowers scientific research and gain inspiration in their own scientific field.

Cognitive diagnosis seeks to estimate the cognitive states of students by exploring their logged practice quiz data. It plays a pivotal role in personalized learning guidance within intelligent education systems. In this paper, we focus on an important, practical, yet often underexplored task: domain-level zero-shot cognitive diagnosis (DZCD), which arises due to the absence of student practice logs in newly launched domains. Recent cross-domain diagnostic models have been demonstrated to be a promising strategy for DZCD. These methods primarily focus on how to transfer student states across domains. However, they might inadvertently incorporate non-transferable information into student representations, thereby limiting the efficacy of knowledge transfer. To tackle this, we propose Zero-1-to-3, a domain-level zero-shot cognitive diagnosis framework via one batch of early-bird students towards three diagnostic objectives. Our approach initiates with pre-training a diagnosis model with dual regularizers, which decouples student states into domain-shared and domain-specific parts. The shared cognitive signals can be transferred to the target domain, enriching the cognitive priors for the new domain, which ensures the cognitive state propagation objective. Subsequently, we devise a strategy to generate simulated practice logs for cold-start students through analyzing the behavioral patterns from early-bird students, fulfilling the domain-adaption goal. Consequently, we refine the cognitive states of cold-start students as diagnostic outcomes via virtual data, aligning with the diagnosis-oriented goal. Finally, extensive experiments on six real-world datasets highlight the efficacy of our model for DZCD and its practical application in question recommendation. The code is publicly available at https://github.com/bigdata-ustc/Zero-1-to-3.

Adaptive Testing System (ATS) is a promising testing mode, extensively utilized in standardized tests like the GRE. It offers personalized ability assessment by dynamically adjusting questions based on individual ability levels. Compared to traditional exams, ATS can improve the accuracy of ability estimates while simultaneously reducing the number of questions required. Despite the diverse testing formats of ATS, tailored to different adaptability requirements in various testing scenarios, there is a notable absence of a unified framework for modeling them. In this paper, we introduce a unified data-driven ATS framework that conceptualizes the various testing formats as a hierarchical test structure search problem. It can learn directly from data to solve for the optimal questions for each student, eliminating the need for manual test design. The proposed solution algorithm comes with theoretical guarantees for estimation error and convergence. Empirical results show that our framework maintains assessment accuracy while reducing question count by 20% on average and improving training stability.

Learner modeling is a crucial task in online learning environments, where Cognitive Diagnosis Models (CDMs) are employed to assess learners' knowledge mastery levels based on recorded response logs. However, the prevalence of noise in recorded response data poses significant challenges, including various behaviors such as guess and slip, casual answers, and system-induced errors. The existence of noise degrades the accuracy of diagnosis results and learner performance predictions. In this work, we propose a general framework, Adaptive Response Denoising (AdaRD), designed to salvage CDMs from the influence of noisy learner-exercise responses. AdaRD extends existing CDMs, incorporating primary training for denoised CDMs and auxiliary training for additional denoising support. The primary training employs binary Generalized Cross Entropy (GCE) loss to slow down the large update of learner knowledge states caused by noisy responses. Simultaneously, we utilize the variance of diagnosed knowledge mastery levels between primary and auxiliary diagnosis modules as a criterion to downweight high-variance responses that are likely to be noisy. In this manner, the proposed framework can prune noisy response learning during training, thereby enhancing the accuracy and robustness of CDMs. Extensive experiments on both real-world and synthetic datasets validate AdaRD's effectiveness in mitigating the impact of noisy learner-exercise responses.

In this paper, we introduce DeAR (_Decompose-Analyze-Rethink_), a framework that iteratively builds a reasoning tree to tackle intricate problems within a single large language model (LLM). Unlike approaches that extend or search for rationales, DeAR is featured by 1) adopting a tree-based question decomposition manner to plan the organization of rationales, which mimics the logical planning inherent in human cognition; 2) globally updating the rationales at each reasoning step through natural language feedback. Specifically, the _Decompose_ stage decomposes the question into simpler sub-questions, storing them as new nodes; the _Analyze_ stage generates and self-checks rationales for sub-questions at each node evel; and the _Rethink_ stage updates parent-node rationales based on feedback from their child nodes. By generating and updating the reasoning process from a more global perspective, DeAR constructs more adaptive and accurate logical structures for complex problems, facilitating timely error correction compared to rationale-extension and search-based approaches such as Tree-of-Thoughts (ToT) and Graph-of-Thoughts (GoT). We conduct extensive experiments on three reasoning benchmarks, including ScienceQA, StrategyQA, and GSM8K, which cover a variety of reasoning tasks, demonstrating that our approach significantly reduces logical errors and enhances performance across various LLMs. Furthermore, we validate that DeAR is an efficient method that achieves a superior trade-off between accuracy and reasoning time compared to ToT and GoT.

Automatically solving math word problems (MWPs) is a challenging task for artificial intelligence (AI) and machine learning (ML) research, which aims to answer the problem with a mathematical expression. Many existing solutions simply model the MWP as a sequence of words, which is far from precise solving. To this end, we turn to how humans solve MWPs. Humans read the problem part-by-part and capture dependencies between words for a thorough understanding and infer the expression precisely in a goal-driven manner with knowledge. Moreover, humans can associate different MWPs to help solve the target with related experience. In this article, we present a focused study on an MWP solver by imitating such procedure. Specifically, we first propose a novel hierarchical math solver (HMS) to exploit semantics in one MWP. First, to imitate human reading habits, we propose a novel encoder to learn the semantics guided by dependencies between words following a hierarchical "word-clause-problem" paradigm. Next, we develop a goal-driven tree-based decoder with knowledge application to generate the expression. One step further, to imitate human associating different MWPs for related experience in problem-solving, we extend HMS to the Relation-enHanced Math Solver (RHMS) to utilize the relation between MWPs. First, to capture the structural similarity relation, we develop a meta-structure tool to measure the similarity based on the logical structure of MWPs and construct a graph to associate related MWPs. Then, based on the graph, we learn an improved solver to exploit related experience for higher accuracy and robustness. Finally, we conduct extensive experiments on two large datasets, which demonstrates the effectiveness of the two proposed methods and the superiority of RHMS.

Computerized Adaptive Testing (CAT) arises as a promising personalized test mode in online education, targeting at revealing students' latent knowledge state by selecting test items adaptively. The item selection strategy is the core component of CAT, which searches for the best suitable test item based on students' current estimated ability at each test step. However, existing selection strategies behave in a brute-force manner, which results in the time complexity being linear to the number of items (N) in the item pool, i.e., O(N). Thus, in reality, the search latency becomes the bottleneck for CAT with a large-scale item pool. To this end, we propose a Search-Efficient Computerized Adaptive Testing framework (SECAT), which aims at enhancing CAT with an efficient selection strategy. Specifically, SECAT contains two main phases: item pool indexing and item search. In the item pool indexing phase, we apply a student-aware spatial partition method on the item pool to divide the test items into many sub-spaces, considering the adaptability of test items. In the item search phase, we optimize the traditional single-round search strategy with the asymptotic theory and propose a multi-round search strategy that can further improve the time efficiency. Compared with existing strategies, the time complexity of SECAT decreases from O(N) to O(logN). Across two real-world datasets, SECAT achieves over 200x speed up with negligible accuracy degradation.

This paper describes our submission to the ICASSP 2023 Signal Processing Grand Challenge (SPGC), which focuses on multilingual Alzheimer’s disease (AD) recognition through spontaneous speech. Our approaches include using a variety of acoustic features and silence-related information for AD detection and mini-mental state examination (MMSE) score prediction, and fine-tuning wav2vec2.0 models on speech in various frequency bands for AD detection. Our overall results on the test data outperform the baseline provided by the organizers, achieving 73.9% accuracy in AD detection by fine-tuning our bilingual wav2vec2.0 pre-trained model on the 0-1000Hz frequency band speech, and 4.610 RMSE (r = 0.565) in MMSE prediction through the fusion of eGeMAPS and silence features.

Computerized adaptive testing (CAT), as a tool that can efficiently measure student's ability, has been widely used in various standardized tests (e.g., GMAT and GRE). The adaptivity of CAT refers to the selection of the most informative questions for each student, reducing test length. Existing CAT methods do not explicitly target ability estimation accuracy since there is no student's true ability as ground truth; therefore, these methods cannot be guaranteed to make the estimate converge to the true with such limited responses. In this paper, we analyze the statistical properties of estimation and find a theoretical approximation of the true ability: the ability estimated by full responses to question bank. Based on this, a Bounded Ability Estimation framework for CAT (BECAT) is proposed in a data-summary manner, which selects a question subset that closely matches the gradient of the full responses. Thus, we develop an expected gradient difference approximation to design a simple greedy selection algorithm, and show the rigorous theoretical and error upper-bound guarantees of its ability estimate. Experiments on both real-world and synthetic datasets, show that it can reach the same estimation accuracy using 15\% less questions on average, significantly reducing test length.

Knowledge Tracing (KT) aims at tracing the evolution of the knowledge states along the learning process of a learner. It has become a crucial task for online learning systems to model the learning process of their users, and further provide their users a personalized learning guidance. However, recent developments in KT based on deep neural networks mostly focus on increasing the accuracy of predicting the next performance of students. We argue that current KT modeling, as well as training paradigm, can lead to models tracing patterns of learner’s learning activities, instead of their evolving knowledge states. In this paper, we propose a new architecture, Diagnostic Transformer (DTransformer), along with a new training paradigm, to tackle this challenge. With DTransformer, we build the architecture from question-level to knowledge-level, explicitly diagnosing learner’s knowledge proficiency from each question mastery states. We also propose a novel training algorithm based on contrastive learning that focuses on maintaining the stability of the knowledge state diagnosis. Through extensive experiments, we will show that with its understanding of knowledge state evolution, DTransformer achieves a better performance prediction accuracy and more stable knowledge state tracing results. We will also show that DTransformer is less sensitive to specific patterns with case study. We open-sourced our code and data at https://github.com/yxonic/DTransformer.

This paper proposes a method of detecting dementia from the simultaneous speech and eye-tracking recordings of subjects in a picture description task. First, automatic speech recognition (ASR) and regional picture recognition (RPR) models are built to extract content-related bottleneck (BN) features for both speech and eye-tracking inputs. Then, a neural network is designed to fuse these two modals for discriminating dementia patients from healthy controls. The network contains a cross-modal Transformer encoder for bimodal interaction, and a self-attention Transformer encoder for final classification. Experimental results demonstrate that the detection accuracy of the proposed method is 84.26%, which outperforms baseline methods and ablated models using single speech or eye-tracking input.

Longitudinal cohort study is an effective way to probe into the risk factors of disease and evaluate intervention measures. It has gradually become the mainstream research method in precision medicine, chronic disease management and evidence-based education, and has been deployed in many National Science and Technology Major Projects, which established its specialized cohorts for natural populations, chronic diseases, specialized diseases. The quality of data is a make-or-break factor of longitudinal cohort study. The subjects and test tasks in longitudinal cohort study have dynamic changes over time, and the data generated involves multiple modalities and scales. Therefore, exploring how to model business-oriented longitudinal cohort data will contribute to a unified understanding and governance of longitudinal cohort data, and ultimately improve data quality. On the one hand, because different modal data in longitudinal cohort study have different dimensional indicators, it is difficult to carry out data modeling based on unified dimensional indicators through simple dimensional splicing; on the other hand, the needs of the longitudinal cohort management scenario determine the calculations should be focused on the granularity of individual subjects and data modal types. Considering the above, the traditional multi-dimensional data modeling method based on data dimension indicators and their measurements as basic elements couldn’t be fully adapted to the counting and statistical requirements under the longitudinal cohort scenarios. This paper proposes a data cube model based on MOLAP named SubTaP, which take multimodal data objects as basic granularity. This model constructs a cube structure with three dimensions of Subject, Task and Phase. It can be applied to meet the visualization requirements of longitudinal cohort management scenario and guide the construction of a data information platform for cohort study. At the same time, it helps to build a unified understanding of longitudinal cohort study data among data generators, cohort maintainers, and data users.

Alzheimer’s disease is a chronic neurodegenerative disease that usually starts slowly and gradually worsens over time. Although there’s no cure for Alzheimer’s disease yet, a number of recent researches have shown that the early diagnosis and intervention could not only improve the quality of life but also help to slow the progression of the disease. Clock Drawing Test (CDT) is one of the commonly used clinical methods for screening cognitive impairment, due to its simplicity and convenience. In this paper, we’d like to introduce DCDT, a novel Clock Drawing Test system based on digital collection and intellectualized analysis. We first introduce the background of AD and CDT, and then describe the DCDT system from the external and internal aspects. Finally, the demonstration scenario is described briefly.

With the pervasive use of GPS-enabled smart phones, location-based services, e.g., Location Based Social Networking (LBSN) have emerged . Point-of-Interests (POIs) Recommendation, as a typical component in LBSN, provides additional values to both customers and merchants in terms of user experience and business turnover. Existing POI recommendation systems mainly adopt Collaborative Filtering (CF), which only exploits user given ratings (i.e., user overall evaluation) about a merchant while regardless of the user preference difference across multiple aspects, which exists commonly in real scenarios. Meanwhile, besides ratings, most LBSNs also provide the review function to allow customers to give their opinions when dealing with merchants, which is often overlooked in these recommender systems. In this demo, we present MARS, a novel POI recommender system based on multi-aspect user preference learning from reviews by using utility theory. We first introduce the organization of our system, and then show how the user preferences across multiple aspects are integrated into our system alongside several case studies of mining user preference and POI recommendations.

With the pervasive use of mobile devices, Location Based Social Networks(LBSNs) have emerged in past years. These LBSNs, allowing their users to share personal experiences and opinions on visited merchants, have very rich and useful information which enables a new breed of location-based services, namely, Merchant Recommendation. Existing techniques for merchant recommendation simply treat each merchant as an item and apply conventional recommendation algorithms, e.g., Collaborative Filtering, to recommend merchants to a target user. However, they do not differentiate the user's real preferences on various aspects, and thus can only achieve limited success. In this paper, we aim to address this problem by utilizing and analyzing user reviews to discover user preferences in different aspects. Following the intuition that a user rating represents a personalized rational choice, we propose a novel utility-based approach by combining collaborative and individual views to estimate user preference (i.e., rating). An optimization algorithm based on a Gaussian model is developed to train our merchant recommendation approach. Lastly we evaluate the proposed approach in terms of effectiveness, efficiency and cold-start using two real-world datasets. The experimental results show that our approach outperforms the state-of-the-art methods. Meanwhile, a real mobile application is implemented to demonstrate the practicability of our method.

With lots of applications emerging in World Wide Web, many interaction data from users are collected and exploited to discover user behavior or interest patterns. In this paper, we attempt to exploit a new interaction data, namely print logs, where each record is printing URLs selected by a user using a popular web printing tool. Users usually print web contents based on an intention (subtask or task). Apparently, mining common print tasks from print logs is able to capture users' intentions, which undoubtedly benefits many web applications, such as task oriented recommendation and behavior targeting. However, it is not an easy job to perform this due to the difficulty of URL topic representation and task formulation. To this end, we propose a general framework, named UPT (Users Print Tasks mining framework), for mining print tasks from print logs. Specifically, we attempt to leverage delicious (a social book marking web service) as an external thesaurus to expand the expression of each URL by selecting tags associated with the domain of each URL. Then, we construct a tag co-occurrence graph where similar tags can be clustered as subtasks. If we view each subtask as an item, then the print log is transformed to a transaction database, on which an efficient pattern mining algorithm is proposed to induce tasks. Finally, we evaluate the effectiveness of the proposed framework through experiments on a real print log.

Abstract A method of learning adaptation rules for case‐based reasoning (CBR) is proposed in this paper. The resource space model and the semantic link network are applied in case‐base construction for efficient resource management and reuse. Adaptation rules are generated from the case‐base with the guidance of domain knowledge, which is also extracted from the case‐base. The adaptation rules are refined before they are applied in the revision process. General domain knowledge is brought in to help accurate similarity computing. After solving each new problem, the adaptation rule set is updated by an evolution module in the retention process. The results of our experiment show that the obtained adaptation rules can improve the performance of the CBR system compared with a retrieval‐only CBR system. The average solution difference error is decreased by 46.56%. Copyright © 2008 John Wiley & Sons, Ltd.