With the rapid advancement and strong generalization capabilities of large language models (LLMs), they have been increasingly incorporated into the active learning pipelines as annotators to reduce annotation costs. However, considering the annotation quality, labels generated by LLMs often fall short of real-world applicability. To address this, we propose a novel active learning framework, Mixture of LLMs in the Loop Active Learning, replacing human annotators with labels generated through a Mixture-of-LLMs-based annotation model, aimed at enhancing LLM-based annotation robustness by aggregating the strengths of multiple LLMs. To further mitigate the impact of the noisy labels, we introduce annotation discrepancy and negative learning to identify the unreliable annotations and enhance learning effectiveness. Extensive experiments demonstrate that our framework achieves performance comparable to human annotation and consistently outperforms single-LLM baselines and other LLM-ensemble-based approaches. Moreover, our framework is built on lightweight LLMs, enabling it to operate fully on local machines in real-world applications.

Recently, automated Interictal Epileptiform Discharge (IED) detection has attracted significant attention as a challenging predictive data analysis task aimed at improving early epilepsy diagnosis. Automated IED detection simplifies visual inspection and assists clinicians in identifying crucial IED waveform patterns in Electroencephalographic (EEG) brain activities. However, IEDs are vastly outnumbered by non-IED or background data, directly training on such data leads to a detrimental impact on model performance. Moreover, most existing methods lack high precision when tested on cross-institution datasets and this will lead to time wasted by the neurologist having to look at predicted IEDs that are not IEDs. To address these issues, we propose a novel approach that employs probabilistic diffusion models for data augmentation alongside Area Under the Precision–Recall Curve (AUPRC) maximization methods. This approach effectively addresses data scarcity and imbalance by combining real and synthesized IEDs to fully balance the training dataset, resulting in a 4.5% increase in precision and a 0.7% improvement in the F1-score during within-data evaluation. Additionally, it proves to be robust and generalizable, as evidenced by a 40.04% and 18.74% enhancement in precision and F1-score when applied to cross-data evaluation using data from another hospital.

Session-based recommendation (SBR) in service computing is pivotal in predicting a user's next action based on their current anonymous session. While Graph Neural Network (GNN)-based methods have shown promise in capturing intricate item transformation relationships within sessions, they often fall short in accurately modeling user preferences. This is primarily due to the common practice of solely considering the last item in the session as the user's current interest, neglecting potentially valuable information embedded in other session items which is essential for capturing user global preferences. Moreover, existing models typically optimize performance solely through cross-entropy loss between predicted items and ground truth labels, while overlooking latent valuable knowledge embedded in intermediate features and item-item relationships that lends support to the model in accurately capturing and modeling user preferences. To address these shortcomings, we propose Multi-Scale Collaborative Distillation (MSCD) for SBR. Our approach introduces a current interest adaptive selection module, which dynamically selects appropriate item embeddings as session-local embeddings by evaluating the importance of each item within the session. This allows for a more accurate capture of the user's current true preferences. Additionally, we propose collaborative knowledge distillation, where multiple models are trained concurrently, enabling the transfer of three types of knowledge including response-based, feature-based, and relationship-based knowledge between models, thereby enriching the model's understanding of user preferences. Experimental evaluations conducted on three popular SBR datasets demonstrate that our MSCD model outperforms recent state-of-the-art methods in terms of recommendation accuracy. Our codes are available at:https://github.com/lonely-ice/MSCD.

Topic modeling is a fundamental task in natural language processing, allowing the discovery of latent thematic structures in text corpora.While Large Language Models (LLMs) have demonstrated promising capabilities in topic discovery, their direct application to topic modeling suffers from issues such as incomplete topic coverage, misalignment of topics, and inefficiency.To address these limitations, we propose LLM-ITL, a novel LLM-in-theloop framework that integrates LLMs with Neural Topic Models (NTMs).In LLM-ITL, global topics and document representations are learned through the NTM.Meanwhile, an LLM refines these topics using an Optimal Transport (OT)-based alignment objective, where the refinement is dynamically adjusted based on the LLM's confidence in suggesting topical words for each set of input words.With the flexibility of being integrated into many existing NTMs, the proposed approach enhances the interpretability of topics while preserving the efficiency of NTMs in learning topics and document representations.Extensive experiments demonstrate that LLM-ITL helps NTMs significantly improve their topic interpretability while maintaining the quality of document representation.

Semi-supervised learning (SSL) has garnered significant attention due to its ability to leverage limited labeled data and a large amount of unlabeled data to improve model generalization performance. Recent approaches achieve impressive successes by combining ideas from both consistency regularization and pseudo-labeling. However, these methods tend to underperform in the more realistic situations with relatively scarce labeled data. We argue that this issue arises because existing methods rely solely on the model’s confidence, making them challenging to accurately assess the model’s state and identify unlabeled examples contributing to the training phase when supervision information is limited, especially during the early stages of model training. In this paper, we propose a novel SSL model called CGMatch, which, for the first time, incorporates a new metric known as Count-Gap (CG). We demonstrate that CG is effective in discovering unlabeled examples beneficial for model training. Along with confidence, a commonly used metric in SSL, we propose a fine-grained dynamic selection (FDS) strategy. This strategy dynamically divides the unlabeled dataset into three subsets with different characteristics: easy-to-learn set, ambiguous set, and hard-to-learn set. By selective filtering subsets, and applying corresponding regularization with selected subsets, we mitigate the negative impact of incorrect pseudo-labels on model optimization and generalization. Extensive experimental results on several common SSL benchmarks indicate the effectiveness of CGMatch especially when the labeled data are particularly limited. Source code is available at https://github.com/BoCheng-96/CGMatch.

Resolving conflicts is critical for improving the reliability of multi-view classification. While prior work focuses on learning consistent and informative representations across views, it often assumes perfect alignment and equal importance of all views, an assumption rarely met in real-world scenarios, as some views may express distinct information. To address this, we develop a computational trust-based discounting method that enhances the Evidential Multi-view framework by accounting for the instance-wise reliability of each view through a probability-sensitive trust mechanism. We evaluate our method on six real-world datasets using Top-1 Accuracy, Fleiss’ Kappa, and a new metric, Multi-View Agreement with Ground Truth, to assess prediction reliability. We also assess the effectiveness of uncertainty in indicating prediction correctness via AUROC. Additionally, we test the scalability of our method through end-to-end training on a large-scale dataset. The experimental results show that computational trust can effectively resolve conflicts, paving the way for more reliable multi-view classification models in real-world applications. Codes available at: https://github.com/OverfitFlow/Trust4Conflict

Existing multi-view learning models struggle in open-set scenarios due to their implicit assumption of class completeness. Moreover, static view-induced biases, which arise from spurious view-label associations formed during training, further degrade their ability to recognize unknown categories. In this paper, we propose a multi-view open-set learning framework via ambiguity uncertainty calibration and view-wise debiasing. To simulate ambiguous samples, we design O-Mix, a novel synthesis strategy to generate virtual samples with calibrated open-set ambiguity uncertainty. These samples are further processed by an auxiliary ambiguity perception network that captures atypical patterns for improved open-set adaptation. Furthermore, we incorporate an HSIC-based contrastive debiasing module that enforces independence between view-specific ambiguous and view-consistent representations, encouraging the model to learn generalizable features. Extensive experiments on diverse multi-view benchmarks demonstrate that the proposed framework consistently enhances unknown-class recognition while preserving strong closed-set performance. The source code are available https://github.com/ZihanFang11/2025_MOCD_ACMMM

The primary challenge of multi-label active learning, differing it from multi-class active learning, lies in assessing the informativeness of an indefinite number of labels while also accounting for the inherited label correlation. Existing studies either require substantial computational resources to leverage correlations or fail to fully explore label dependencies. Additionally, real-world scenarios often require addressing intrinsic biases stemming from imbalanced data distributions. In this paper, we propose a new multi-label active learning strategy to address both challenges. Our method incorporates progressively updated positive and negative correlation matrices to capture co-occurrence and disjoint relationships within the label space of annotated samples, enabling a holistic assessment of uncertainty rather than treating labels as isolated elements. Furthermore, alongside diversity, our model employs ensemble pseudo labeling and beta scoring rules to address data imbalances. Extensive experiments on four realistic datasets demonstrate that our strategy consistently achieves more reliable and superior performance, compared to several established methods.

Abstract Topic modeling has been a widely used tool for unsupervised text analysis. However, comprehensive evaluations of a topic model remain challenging. Existing evaluation methods are either less comparable across different models (e.g., perplexity) or focus on only one specific aspect of a model (e.g., topic quality or document representation quality) at a time, which is insufficient to reflect the overall model performance. In this paper, we propose WALM (Word Agreement with Language Model), a new evaluation method for topic modeling that considers the semantic quality of document representations and topics in a joint manner, leveraging the power of Large Language Models (LLMs). With extensive experiments involving different types of topic models, WALM is shown to align with human judgment and can serve as a complementary evaluation method to the existing ones, bringing a new perspective to topic modeling. Our software package is available at https://github.com/Xiaohao-Yang/Topic_Model_Evaluation.

Within the scope of natural language processing, the domain of multi-label text classification is uniquely challenging due to its expansive and uneven label distribution. The complexity deepens due to the demand for an extensive set of annotated data for training an advanced deep learning model, especially in specialized fields where the labeling task can be labor-intensive and often requires domain-specific knowledge. Addressing these challenges, our study introduces a novel deep active learning strategy, capitalizing on the Beta family of proper scoring rules within the Expected Loss Reduction framework. It computes the expected increase in scores using the Beta Scoring Rules, which are then transformed into sample vector representations. These vector representations guide the diverse selection of informative sample, directly linking this process to the model's expected proper score. Comprehensive evaluations across both synthetic and real datasets reveal our method's capability to often outperform established acquisition techniques in multi-label text classification, presenting encouraging outcomes across various architectural and dataset scenarios.

The thermal-to-visible (T2V) face translation task is essential for enabling face verification in low-light or dark conditions by converting thermal infrared faces into their visible counterparts. However, this task faces two primary challenges. First, the inherent differences between the modalities hinder the effective use of thermal information to guide RGB face reconstruction. Second, translated RGB faces often lack the identity details of the corresponding visible faces, such as skin color. To tackle these challenges, we introduce DiffTV, the first Latent Diffusion Model (LDM) specifically designed for T2V facial image translation with a focus on preserving identity. Our approach proposes a novel heterogeneous feature alignment strategy that bridges the modal gap and extracts both coarse-and fine-grained identity features consistent with visible images. Furthermore, a dual-stage condition injection strategy introduces control information to guide identity-preserved translation. Experimental results demonstrate the superior performance of DiffTV, particularly in scenarios where maintaining identity integrity is critical.

Knowledge tracing (KT) aims to leverage students’ learning histories to estimate their mastery levels on a set of pre-defined skills, based on which the corresponding future performance can be accurately predicted. In practice, a student’s learning history comprises answers to sets of massed questions, each known as a session, rather than merely being a sequence of independent answers. Theoretically, within and across these sessions, students’ learning dynamics can be very different. Therefore, how to effectively model the dynamics of students’ knowledge states within and across the sessions is crucial for handling the KT problem. Most existing KT models treat student’s learning records as a single continuing sequence, without capturing the sessional shift of students’ knowledge state. To address the above issue, we propose a novel hierarchical transformer model, named HiTSKT, comprises an interaction(-level) encoder to capture the knowledge a student acquires within a session, and a session(-level) encoder to summarise acquired knowledge across the past sessions. To predict an interaction in the current session, a knowledge retriever integrates the summarised past-session knowledge with the previous interactions’ information into proper knowledge representations. These representations are then used to compute the student’s current knowledge state. Additionally, to model the student’s long-term forgetting behaviour across the sessions, a power-law-decay attention mechanism is designed and deployed in the session encoder, allowing it to emphasize more on the recent sessions. Extensive experiments on four public datasets demonstrate that HiTSKT achieves new state-of-the-art performance on all the datasets compared with seven state-of-the-art KT models.

The effectiveness of active learning largely depends on the sampling efficiency of the acquisition function. Expected Loss Reduction (ELR) focuses on a Bayesian estimate of the reduction in classification error, and more general costs fit in the same framework. We propose Bayesian Estimate of Mean Proper Scores (BEMPS) to estimate the increase in strictly proper scores such as log probability or negative mean square error within this framework. We also prove convergence results for this general class of costs. To facilitate better experimentation with the new acquisition functions, we develop a complementary batch AL algorithm that encourages diversity in the vector of expected changes in scores for unlabeled data. To allow high-performance classifiers, we combine deep ensembles, and dynamic validation set construction on pretrained models, and further speed up the ensemble process with the idea of Monte Carlo Dropout. Extensive experiments on both texts and images show that the use of mean square error and log probability with BEMPS yields robust acquisition functions and well-calibrated classifiers, and consistently outperforms the others tested. The advantages of BEMPS over the others are further supported by a set of qualitative analyses, where we visualise their sampling behaviour using data maps and t-SNE plots.

Knowledge distillation (KD) is a prevalent model compression technique in deep learning, aiming to leverage knowledge from a large teacher model to enhance the training of a smaller student model. It has found success in deploying compact deep models in intelligent applications like intelligent transportation, smart health, and distributed intelligence. Current knowledge distillation methods primarily fall into two categories: offline and online knowledge distillation. Offline methods involve a one-way distillation process, transferring unvaried knowledge from teacher to student, while online methods enable the simultaneous training of multiple peer students. However, existing knowledge distillation methods often face challenges where the student may not fully comprehend the teacher's knowledge due to model capacity gaps, and there might be knowledge incongruence among outputs of multiple students without teacher guidance. To address these issues, we propose a novel reciprocal teacher-student learning inspired by human teaching and examining through forward and feedback knowledge distillation (FFKD). Forward knowledge distillation operates offline, while feedback knowledge distillation follows an online scheme. The rationale is that feedback knowledge distillation enables the pre-trained teacher model to receive feedback from students, allowing the teacher to refine its teaching strategies accordingly. To achieve this, we introduce a new weighting constraint to gauge the extent of students' understanding of the teacher's knowledge, which is then utilized to enhance teaching strategies. Experimental results on five visual recognition datasets demonstrate that the proposed FFKD outperforms current state-of-the-art knowledge distillation methods.

Deep dictionary learning (DDL) shows good performance in visual classification tasks. However, almost all existing DDL methods ignore the locality relationships between the input data representations and the learned dictionary atoms, and learn sub-optimal representations in the feature coding stage, which are less conducive to classification. To this end, we propose a hierarchical locality-aware deep dictionary learning (HILADLE) framework for classification, which can learn locality-constrained dictionaries at different abstract levels through hierarchical dictionary learning. The locality constraints play an important role in learning informative dictionary atoms while preserving the data structure in the original input feature space. Moreover, instead of using an identity activation function like existing DDL methods, we further boost the generalization performance of our HILADLE method with a ReLU activation function to deal with the overfitting issue caused by over-parameterization, inspired by its effectiveness in deep neural networks. Finally, the concatenation of all feature representations learned at different layers is used as input to the final classifier. We demonstrate, through an extensive set of experiments on several benchmark face recognition, image classification, and age estimation datasets, that our method is able to surpass several dictionary learning, deep dictionary learning and deep learning methods.

The application of Auto-Encoder (AE) to multi-view representation learning has gained traction due to advancements in deep learning. While some current AE-based multi-view representation learning algorithms incorporate the geometric structure of the input data into their feature representation learning process, their use of a shallow structured graph regularization term can be restrictive when used in conjunction with deep models. Furthermore, current multi-view representation learning algorithms do not fully utilize the diversity and consistency presented in different views, leading to a reduction in the efficacy of feature learning. This paper introduces a novel approach, reconstructed graph constrained auto-encoders (RGCAE), for multi-view representation learning. Unlike existing methods, our approach incorporates deep adaptive graph regularization based on multi-layer perceptron to ensure the preservation of the geometric similarity graph, which is constructed based on the local invariance principle. By decoupling the feature representation learning from the preservation of the geometric structure among different views, our approach can better leverage the diversity presented in multi-view data. We obtain view-specific representations that preserve the geometric structure and then combine them by averaging to obtain a common representation. To ensure the consistency of the multi-view data, we minimize the loss between the view-specific and common representations. Consequently, our RGCAE approach can maintain the geometric structure of multi-view data and is better suited for integration with deep models. Extensive experiments on six datasets demonstrate that RGCAE obtained promising performance, compared with the state-of-the-art methods.

In recent years, various distillation methods for semantic segmentation have been proposed. However, these methods typically train the student model to imitate the intermediate features or logits of the teacher model directly, thereby overlooking the high-discrepancy regions learned by both models, particularly the differences in instance edges. In this paper, we introduce a novel approach, called Difference-aware Distillation, to address this limitation. Our proposed method detects the discrepancies among the teacher model and the student model in the logit space through two masking mechanisms (i.e., masking by logit differences with respect to the ground truth labels and masking by differences in the predictive class probabilities), and guides the student model to restore the teacher's features with the focus on these highly-discrepant regions, resulting in improved segmentation performance. With the features jointly masked by these two mechanisms, the student model learns to preserve the teacher's features via a feature generation module, thus achieving better representation. Our experimental evaluation on three datasets, Cityscapes, Pascal2012, and ADE20 K, demonstrates our proposed approach outperforms several baselines considered. Further visualization analysis confirms that our method effectively directs the student model's attention to the discrepancies, such as the edges of small objects and the interiors of large objects.

Knowledge distillation (KD), as an efficient and effective model compression technique, has received considerable attention in deep learning. The key to its success is about transferring knowledge from a large teacher network to a small student network. However, most existing KD methods consider only one type of knowledge learned from either instance features or relations via a specific distillation strategy, failing to explore the idea of transferring different types of knowledge with different distillation strategies. Moreover, the widely used offline distillation also suffers from a limited learning capacity due to the fixed large-to-small teacher-student architecture. In this article, we devise a collaborative KD via multiknowledge transfer (CKD-MKT) that prompts both self-learning and collaborative learning in a unified framework. Specifically, CKD-MKT utilizes a multiple knowledge transfer framework that assembles self and online distillation strategies to effectively: 1) fuse different kinds of knowledge, which allows multiple students to learn knowledge from both individual instances and instance relations, and 2) guide each other by learning from themselves using collaborative and self-learning. Experiments and ablation studies on six image datasets demonstrate that the proposed CKD-MKT significantly outperforms recent state-of-the-art methods for KD.

Continual learning (CL) is a machine learning paradigm that accumulates knowledge while learning sequentially. The main challenge in CL is catastrophic forgetting of previously seen tasks, which occurs due to shifts in the probability distribution. To retain knowledge, existing CL models often save some past examples and revisit them while learning new tasks. As a result, the size of saved samples dramatically increases as more samples are seen. To address this issue, we introduce an efficient CL method by storing only a few samples to achieve good performance. Specifically, we propose a dynamic prototype-guided memory replay (PMR) module, where synthetic prototypes serve as knowledge representations and guide the sample selection for memory replay. This module is integrated into an online meta-learning (OML) model for efficient knowledge transfer. We conduct extensive experiments on the CL benchmark text classification datasets and examine the effect of training set order on the performance of CL models. The experimental results demonstrate the superiority our approach in terms of accuracy and efficiency.

Current work in named entity recognition (NER) uses either cross entropy (CE) or conditional random fields (CRF) as the objective/loss functions to optimize the underlying NER model. Both of these traditional objective functions for the NER problem generally produce adequate performance when the data distribution is balanced and there are sufficient annotated training examples. But since NER is inherently an imbalanced tagging problem, the model performance under the low-resource settings could suffer using these standard objective functions. Based on recent advances in area under the ROC curve (AUC) maximization, we propose to optimize the NER model by maximizing the AUC score. We give evidence that by simply combining two binary-classifiers that maximize the AUC score, significant performance improvement over traditional loss functions is achieved under low-resource NER settings. We also conduct extensive experiments to demonstrate the advantages of our method under the low-resource and highly-imbalanced data distribution settings. To the best of our knowledge, this is the first work that brings AUC maximization to the NER setting. Furthermore, we show that our method is agnostic to different types of NER embeddings, models and domains. The code of this work is available at https://github.com/dngu0061/NER-AUC-2T.

Named entity recognition (NER), a task that identifies and categorizes named entities such as persons or organizations from text, is traditionally framed as a multi-class classification problem. However, this approach often overlooks the issues of imbalanced label distributions, particularly in low-resource settings, which is common in certain NER contexts, like biomedical NER (bioNER). To address these issues, we propose an innovative reformulation of the multi-class problem as a one-vs-all (OVA) learning problem and introduce a loss function based on the area under the receiver operating characteristic curve (AUC). To enhance the efficiency of our OVA-based approach, we propose two training strategies: one groups labels with similar linguistic characteristics, and another employs meta-learning. The superiority of our approach is confirmed by its performance, which surpasses traditional NER learning in varying NER settings.

Uncertainty estimation is an important research area to make deep neural networks (DNNs) more trustworthy. While extensive research on uncertainty estimation has been conducted with unimodal data, uncertainty estimation for multimodal data remains a challenge. Neural processes (NPs) have been demonstrated to be an effective uncertainty estimation method for unimodal data by providing the reliability of Gaussian processes with efficient and powerful DNNs. While NPs hold significant potential for multimodal uncertainty estimation, the adaptation of NPs for multimodal data has not been carefully studied. To bridge this gap, we propose Multimodal Neural Processes (MNPs) by generalising NPs for multimodal uncertainty estimation. Based on the framework of NPs, MNPs consist of several novel and principled mechanisms tailored to the characteristics of multimodal data. In extensive empirical evaluation, our method achieves state-of-the-art multimodal uncertainty estimation performance, showing its appealing robustness against noisy samples and reliability in out-of-distribution detection with faster computation time compared to the current state-of-the-art multimodal uncertainty estimation method.

Abstract Knowledge distillation is a simple yet effective technique for deep model compression, which aims to transfer the knowledge learned by a large teacher model to a small student model. To mimic how the teacher teaches the student, existing knowledge distillation methods mainly adapt an unidirectional knowledge transfer, where the knowledge extracted from different intermedicate layers of the teacher model is used to guide the student model. However, it turns out that the students can learn more effectively through multi-stage learning with a self-reflection in the real-world education scenario, which is nevertheless ignored by current knowledge distillation methods. Inspired by this, we devise a new knowledge distillation framework entitled multi-target knowledge distillation via student self-reflection or MTKD-SSR, which can not only enhance the teacher’s ability in unfolding the knowledge to be distilled, but also improve the student’s capacity of digesting the knowledge. Specifically, the proposed framework consists of three target knowledge distillation mechanisms: a stage-wise channel distillation (SCD), a stage-wise response distillation (SRD), and a cross-stage review distillation (CRD), where SCD and SRD transfer feature-based knowledge (i.e., channel features) and response-based knowledge (i.e., logits) at different stages, respectively; and CRD encourages the student model to conduct self-reflective learning after each stage by a self-distillation of the response-based knowledge. Experimental results on five popular visual recognition datasets, CIFAR-100, Market-1501, CUB200-2011, ImageNet, and Pascal VOC, demonstrate that the proposed framework significantly outperforms recent state-of-the-art knowledge distillation methods.

Continual learning (CL) aims to enhance sequential learning by alleviating the forgetting of previously acquired knowledge. Recent advances in CL lack consideration of the real-world scenarios, where labeled data are scarce and unlabeled data are abundant. To narrow this gap, we focus on semi-supervised continual learning (SSCL). We exploit unlabeled data under limited supervision in the CL setting and demonstrate the feasibility of semi-supervised learning in CL. In this work, we propose a novel method, namely Meta-SSCL, which combines meta-learning with pseudo-labeling and data augmentations to learn a sequence of semi-supervised tasks without catastrophic forgetting. Extensive experiments on CL benchmark text classification datasets show that our method achieves promising results in SSCL.

Domain adaptation is an effective solution to data scarcity in low-resource scenarios.However, when applied to token-level tasks such as bioNER, domain adaptation methods often suffer from the challenging linguistic characteristics that clinical narratives possess, which leads to unsatsifactory performance.In this paper, we present a simple yet effective hardnessguided domain adaptation (HGDA) framework for bioNER tasks that can effectively leverage the domain hardness information to improve the adaptability of the learnt model in the low-resource scenarios.Experimental results on biomedical datasets show that our model can achieve significant performance improvement over the recently published state-of-theart (SOTA) MetaNER model.

Recently, discrete latent variable models have received a surge of interest in both Natural Language Processing (NLP) and Computer Vision (CV), attributed to their comparable performance to the continuous counterparts in representation learning, while being more interpretable in their predictions.In this paper, we develop a topic-informed discrete latent variable model for semantic textual similarity, which learns a shared latent space for sentence-pair representation via vector quantization.Compared with previous models limited to local semantic contexts, our model can explore richer semantic information via topic modeling.We further boost the performance of semantic similarity by injecting the quantized representation into a transformer-based language model with a well-designed semanticdriven attention mechanism.We demonstrate, through extensive experiments across various English language datasets, that our model is able to surpass several strong neural baselines in semantic textual similarity tasks.

Uncertainty estimation is essential to make neural networks trustworthy in real-world applications. Extensive research efforts have been made to quantify and reduce predictive uncertainty. However, most existing works are designed for unimodal data, whereas multi-view uncertainty estimation has not been sufficiently investigated. Therefore, we propose a new multi-view classification framework for better uncertainty estimation and out-of-domain sample detection, where we associate each view with an uncertainty-aware classifier and combine the predictions of all the views in a principled way. The experimental results with real-world datasets demonstrate that our proposed approach is an accurate, reliable, and well-calibrated classifier, which predominantly outperforms the multi-view baselines tested in terms of expected calibration error, robustness to noise, and accuracy for the in-domain sample classification and the out-of-domain sample detection tasks

Legal judgment prediction (LJP) is used to predict judgment results based on the description of individual legal cases. In order to be more suitable for actual application scenarios in which the case has cited multiple articles and has multiple charges, we formulate legal judgment prediction as a multiple label learning problem and present a deep learning model that can effectively encode the content of each legal case via a multi-residual convolution neural network and the semantics of law articles via an article encoder. An article-wise attention mechanism is proposed to fuse the two types of encoded information. Experimental results derived on the CAIL2018 datasets show that our model provides a significant performance improvement over the existing neural models in predicting relevant law articles and charges.

Recently, deep dictionary learning (DDL) has aroused attention due to its abilities of learning multiple different dictionaries and extracting multi-level abstract feature representations for samples. It has been applied to many intelligent recognition tasks, such as vehicle detection, traffic sign recognition and driver monitoring. Nevertheless, the off-the-shelf DDL-based methods ignore the essential structural information of data in multi-layer dictionary learning. The learned hierarchical data representations are less discriminative. To address this issue, we develop a new DDL framework, called the hierarchical graph augmented deep collaborative dictionary learning (HGDCDL). Firstly, we propose a new deep collaborative dictionary learning (DCDL) that applies collaborative representation to the deepest-level representation learning. Most importantly, equipped with a simple yet effective hierarchal graph construction mechanism, our HGDCDL uses the structure of data to regularize dictionary learning, and generates more informative dictionaries and discriminative representations at different levels. Extensive experiments show that our HGDCDL performs significantly better than the state-of-the-art shallow and deep representation learning methods for classification.

Neural topic models (NTMs) apply deep neural networks to topic modelling.Despite their success, NTMs generally ignore two important aspects: (1) only document-level word count information is utilized for the training, while more fine-grained sentence-level information is ignored, and (2) external semantic knowledge regarding documents, sentences and words are not exploited for the training.To address these issues, we propose a variational autoencoder (VAE) NTM model that jointly reconstructs the sentence and document word counts using combinations of bag-of-words (BoW) topical embeddings and pre-trained semantic embeddings.The pre-trained embeddings are first transformed into a common latent topical space to align their semantics with the BoW embeddings.Our model also features hierarchical KL divergence to leverage embeddings of each document to regularize those of their sentences, thereby paying more attention to semantically relevant sentences.Both quantitative and qualitative experiments have shown the efficacy of our model in 1) lowering the reconstruction errors at both the sentence and document levels, and 2) discovering more coherent topics from real-world datasets.

Topic modelling has been a successful technique for text analysis for almost twenty years. When topic modelling met deep neural networks, there emerged a new and increasingly popular research area, neural topic models, with nearly a hundred models developed and a wide range of applications in neural language understanding such as text generation, summarisation and language models. There is a need to summarise research developments and discuss open problems and future directions. In this paper, we provide a focused yet comprehensive overview of neural topic models for interested researchers in the AI community, so as to facilitate them to navigate and innovate in this fast-growing research area. To the best of our knowledge, ours is the first review on this specific topic.

We study acquisition functions for active learning (AL) for text classification. The Expected Loss Reduction (ELR) method focuses on a Bayesian estimate of the reduction in classification error, recently updated with Mean Objective Cost of Uncertainty (MOCU). We convert the ELR framework to estimate the increase in (strictly proper) scores like log probability or negative mean square error, which we call Bayesian Estimate of Mean Proper Scores (BEMPS). We also prove convergence results borrowing techniques used with MOCU. In order to allow better experimentation with the new acquisition functions, we develop a complementary batch AL algorithm, which encourages diversity in the vector of expected changes in scores for unlabelled data. To allow high performance text classifiers, we combine ensembling and dynamic validation set construction on pretrained language models. Extensive experimental evaluation then explores how these different acquisition functions perform. The results show that the use of mean square error and log probability with BEMPS yields robust acquisition functions, which consistently outperform the others tested.

Representation-based classification (RBC) has been attracting a great deal of attention in pattern recognition. As a typical extension to RBC, collaborative representation-based classification (CRC) has demonstrated its superior performance in various image classification tasks. Ideally, we expect that the learned class-specific representations for a testing sample are discriminative, and the representation computed for the true class dominates the final representation of the testing sample. Most existing CRC-based methods can learn pattern discrimination, but cannot differentiate the contribution of class-specific representations to the classification of each testing sample. It is challenging for a representation-based classifier to retain both properties. To address this challenge and further improve CRC's classification performance, we propose a novel CRC-based method, class mean-weighted discriminative collaborative representation-based classifier (CMW-DCRC). Its objective function penalises the standard l 2 -norm residuals with two discriminative regularisation terms. A decorrelating term makes the class-specific representations more discriminative, and a newly designed class mean-weighted term that promotes the training samples from individual classes to competitively reconstruct the testing sample while boosting the contribution of the true class. To further enhance the robustness of CRC, we extend CMW-DCRC by replacing the l2-norm coding residual with a l1-norm coding residual, and solve the optimisation problem with an iteratively reweighted least square algorithm. Extensive experimental results on nine image data sets have shown that our methods outperform the state-of-the-art RBC-based methods.

Graph neural networks (GNNs) are important tools for transductive learning tasks, such as node classification in graphs, due to their expressive power in capturing complex interdependency between nodes. To enable GNN learning, existing works typically assume that labeled nodes, from two or multiple classes, are provided, so that a discriminative classifier can be learned from the labeled data. In reality, this assumption might be too restrictive for applications, as users may only provide labels of interest in a single class for a small number of nodes. In addition, most GNN models only aggregate information from short distances ( e.g. , 1-hop neighbors) in each round, and fail to capture long-distance relationship in graphs. In this article, we propose a novel GNN framework, long-short distance aggregation networks, to overcome these limitations. By generating multiple graphs at different distance levels, based on the adjacency matrix, we develop a long-short distance attention model to model these graphs. The direct neighbors are captured via a short-distance attention mechanism, and neighbors with long distance are captured by a long-distance attention mechanism. Two novel risk estimators are further employed to aggregate long-short-distance networks, for PU learning and the loss is back-propagated for model learning. Experimental results on real-world datasets demonstrate the effectiveness of our algorithm.

Many applications, such as text modelling, high-throughput sequencing, and recommender systems, require analysing sparse, high-dimensional, and overdispersed discrete (count or binary) data. Recent deep probabilistic models based on variational autoencoders (VAE) have shown promising results on discrete data but may have inferior modelling performance due to the insufficient capability in modelling overdispersion and model misspecification. To address these issues, we develop a VAE-based framework using the negative binomial distribution as the data distribution. We also provide an analysis of its properties vis-à-vis other models. We conduct extensive experiments on three problems from discrete data analysis: text analysis/topic modelling, collaborative filtering, and multi-label learning. Our models outperform state-of-the-art approaches on these problems, while also capturing the phenomenon of overdispersion more effectively.

Few/Zero-shot learning is a big challenge of many classifications tasks, where a classifier is required to recognise instances of classes that have very few or even no training samples. It becomes more difficult in multilabel classification, where each instance is la-belled with more than one class. In this paper, we present a simple multi-graph aggregation model that fuses knowledge from multiple label graphs encoding different semantic label relationships in order to study how the aggregated knowledge can benefit multi-label zero/few-shot document classification. The model utilises three kinds of semantic information, i.e., the pre-trained word embeddings, label description, and pre-defined label relations. Experimental results derived on two large clinical datasets (i.e., MIMIC-II and MIMIC-III) and the EU legislation dataset show that methods equipped with the multi-graph knowledge aggregation achieve significant performance improvement across almost all the measures on few/zero-shot labels.

Obtaining training data for multi-document Summarization (MDS) is time consuming and resource-intensive, so recent neural models can only be trained for limited domains. In this paper, we propose SummPip: an unsupervised method for multi-document summarization, in which we convert the original documents to a sentence graph, taking both linguistic and deep representation into account, then apply spectral clustering to obtain multiple clusters of sentences, and finally compress each cluster to generate the final summary. Experiments on Multi-News and DUC-2004 datasets show that our method is competitive to previous unsupervised methods and is even comparable to the neural supervised approaches. In addition, human evaluation shows our system produces consistent and complete summaries compared to human written ones.

Analysing topics in short texts (e.g., tweets and new headings) is a challenging task because short texts often contain insufficient word co-occurrence information, which is important to learn good topics in conventional topic topics.To deal with the insufficiency, we propose a generative model that aggregates short texts into clusters by leveraging the associated meta information.Our model can generate more interpretable topics as well as document clusters.We develop an effective Gibbs sampling algorithm favoured by the fully local conjugacy in the model.Extensive experiments demonstrate that our model achieves better performance in terms of document clustering and topic coherence.

Graph neural nets are emerging tools to represent network nodes for classification. However, existing approaches typically suffer from two limitations: (1) they only aggregate information from short distance (e.g., 1-hop neighbors) each round and fail to capturelong distance relationship in graphs; (2) they require users to label data from several classes to facilitate the learning of discriminative models; whereas in reality, users may only provide labels of a small number of nodes in a single class. To overcome these limitations, this paper presents a novel long-short distance aggregation networks (\textttLSDAN ) for positive unlabeled (PU) graph learning. Our theme is to generate multiple graphs at different distances based on the adjacency matrix, and further develop a long-short distance attention model for these graphs. The short-distance attention mechanism is used to capture the importance of neighbor nodes to a target node. The long-distance attention mechanism is used to capture the propagation of information within a localized area of each node and help model weights of different graphs for node representation learning. A non-negative risk estimator is further employed, to aggregate long- short-distance networks, for PU learning using back-propagated loss modeling. Experiments on real-world datasets validate the effectiveness of our approach.

We present a probabilistic, fully Bayesian framework for multi-label learning. Our framework is based on the idea of learning a joint low-rank embedding of the label matrix and the label co-occurrence matrix. The proposed framework has the following appealing aspects: (1) It leverages the sparsity in the label matrix and the feature matrix, which results in very efficient inference, especially for sparse datasets, commonly encountered in multi-label learning problems, and (2) By effectively utilizing the label co-occurrence information, the model yields improved prediction accuracies, especially in the case where the amount of training data is low and/or the label matrix has a significant fraction of missing labels. Our framework enjoys full local conjugacy and admits a simple inference procedure via a scalable Gibbs sampler. We report experimental results on a number of benchmark datasets, on which it outperforms several state-of-the-art multi-label learning models.

One important task of topic modeling for text analysis is interpretability. By discovering structured topics one is able to yield improved interpretability as well as modeling accuracy. In this paper, we propose a novel topic model with a deep structure that explores both inter-topic and intra-topic structures informed by word embeddings. Specifically, our model discovers inter topic structures in the form of topic hierarchies and discovers intra topic structures in the form of sub-topics, each of which is informed by word embeddings and captures a fine-grained thematic aspect of a normal topic. Extensive experiments demonstrate that our model achieves the state-of-the-art performance in terms of perplexity, document classification, and topic quality. Moreover, with topic hierarchies and sub-topics, the topics discovered in our model are more interpretable, providing an illuminating means to understand text data.

Recently, considerable research effort has been devoted to developing deep architectures for topic models to learn topic structures. Although several deep models have been proposed to learn better topic proportions of documents, how to leverage the benefits of deep structures for learning word distributions of topics has not yet been rigorously studied. Here we propose a new multi-layer generative process on word distributions of topics, where each layer consists of a set of topics and each topic is drawn from a mixture of the topics of the layer above. As the topics in all layers can be directly interpreted by words, the proposed model is able to discover interpretable topic hierarchies. As a self-contained module, our model can be flexibly adapted to different kinds of topic models to improve their modelling accuracy and interpretability. Extensive experiments on text corpora demonstrate the advantages of the proposed model.

Most work on segmenting text does so on the basis of topic changes, but it can be of interest to segment by other, stylistically expressed characteristics such as change of authorship or native language.We propose a Bayesian unsupervised text segmentation approach to the latter.While baseline models achieve essentially random segmentation on our task, indicating its difficulty, a Bayesian model that incorporates appropriately compact language models and alternating asymmetric priors can achieve scores on the standard metrics around halfway to perfect segmentation.

In natural language processing and related fields, it has been shown that the word embeddings can successfully capture both the semantic and syntactic features of words. They can serve as complementary information to topics models, especially for the cases where word co-occurrence data is insufficient, such as with short texts. In this paper, we propose a focused topic model where how a topic focuses on words is informed by word embeddings. Our models is able to discover more informed and focused topics with more representative words, leading to better modelling accuracy and topic quality. With the data argumentation technique, we can derive an efficient Gibbs sampling algorithm that benefits from the fully local conjugacy of the model. We conduct extensive experiments on several real world datasets, which demonstrate that our model achieves comparable or improved performance in terms of both perplexity and topic coherence, particularly in handling short text data.

Besides the text content, documents and their associated words usually come with rich sets of meta information, such as categories of documents and semantic/syntactic features of words, like those encoded in word embeddings. Incorporating such meta information directly into the generative process of topic models can improve modelling accuracy and topic quality, especially in the case where the word-occurrence information in the training data is insufficient. In this paper, we present a topic model, called MetaLDA, which is able to leverage either document or word meta information, or both of them jointly. With two data argumentation techniques, we can derive an efficient Gibbs sampling algorithm, which benefits from the fully local conjugacy of the model. Moreover, the algorithm is favoured by the sparsity of the meta information. Extensive experiments on several real world datasets demonstrate that our model achieves comparable or improved performance in terms of both perplexity and topic quality, particularly in handling sparse texts. In addition, compared with other models using meta information, our model runs significantly faster.

Relational data are usually highly incomplete in practice, which inspires us to leverage side information to improve the performance of community detection and link prediction. This paper presents a Bayesian probabilistic approach that incorporates various kinds of node attributes encoded in binary form in relational models with Poisson likelihood. Our method works flexibly with both directed and undirected relational networks. The inference can be done by efficient Gibbs sampling which leverages sparsity of both networks and node attributes. Extensive experiments show that our models achieve the state-of-the-art link prediction results, especially with highly incomplete relational data.

Nowadays, users of social networks like tweets and weibo have generated massive geo-tagged records, and these records reveal their activities in the physical world together with spatio-temporal dynamics. Existing trajectory data management studies mainly focus on analyzing the spatio-temporal properties of trajectories, while leaving the understanding of their activities largely untouched. In this paper, we incorporate the semantic analysis of the activity information embedded in trajectories into query modelling and processing, with the aim of providing end users more accurate and meaningful trip recommendations. To this end, we propose a novel trajectory query that not only considers the spatio-temporal closeness but also, more importantly, leverages probabilistic topic modelling to capture the semantic relevance of the activities between data and query. To support efficient query processing, we design a novel hybrid index structure, namely ST-tree, to organize the trajectory points hierarchically, which enables us to prune the search space in spatial and topic dimensions simultaneously. The experimental results on real datasets demonstrate the efficiency and scalability of the proposed index structure and search algorithms.

In this paper, we propose a new dimensionality reduction method called discriminative sparsity preserving graph embedding (DSPGE). Unlike many existing graph embedding methods such as locality preserving projections (LPP) and sparsity preserving projections (SPP), the aim of DSPGE is to preserve the sparse reconstructive relationships of data while simultaneously capture the geometric and discriminant structure of data in the embedding space. Through the sparse reconstruction and class-specific adjacent graphs, DSPGE characterizes the intra-class and inter-class sparsity preserving scatters, seeking to achieve the optimal projections that simultaneously maximize the inter-class sparsity preserving scatter and minimize intra-class sparsity preserving scatter. The effectiveness of the proposed DSPGE is demonstrated on two popular face databases, compared to up-to-date methods. The experimental results show that DSPGE outperforms the competing methods with the satisfactory classification performance.

Documents from the same domain usually discuss similar topics in a similar order. However, the number of topics and the exact topics discussed in each individual document can vary. In this paper we present a simple topic model that uses generalised Mallows models and incomplete topic orderings to incorporate this ordering regularity into the probabilistic generative process of the new model. We show how to reparameterise the new model so that a point-wise sampling algorithm from the Bayesian word segmentation literature can be used for inference. This algorithm jointly samples not only the topic orders and the topic assignments but also topic segmentations of documents. Experimental results show that our model performs significantly better than the other ordering-based topic models on nearly all the corpora that we used, and competitively with other state-of-the-art topic segmentation models on corpora that have a strong ordering regularity.

Zhendong Zhao, Lan Du, Benjamin Börschinger, John K Pate, Massimiliano Ciaramita, Mark Steedman, Mark Johnson. Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2015.

In applications we may want to compare different document collections: they could have shared content but also different and unique aspects in particular collections. This task has been called comparative text mining or cross-collection modeling. We present a differential topic model for this application that models both topic differences and similarities. For this we use hierarchical Bayesian nonparametric models. Moreover, we found it was important to properly model power-law phenomena in topic-word distributions and thus we used the full Pitman-Yor process rather than just a Dirichlet process. Furthermore, we propose the transformed Pitman-Yor process (TPYP) to incorporate prior knowledge such as vocabulary variations in different collections into the model. To deal with the non-conjugate issue between model prior and likelihood in the TPYP, we thus propose an efficient sampling algorithm using a data augmentation technique based on the multinomial theorem. Experimental results show the model discovers interesting aspects of different collections. We also show the proposed MCMC based algorithm achieves a dramatically reduced test perplexity compared to some existing topic models. Finally, we show our model outperforms the state-of-the-art for document classification/ideology prediction on a number of text collections.

Probabilistic topic models are widely used to discover latent topics in document collections, while latent feature vector representations of words have been used to obtain high performance in many NLP tasks. In this paper, we extend two different Dirichlet multinomial topic models by incorporating latent feature vector representations of words trained on very large corpora to improve the word-topic mapping learnt on a smaller corpus. Experimental results show that by using information from the external corpora, our new models produce significant improvements on topic coherence, document clustering and document classification tasks, especially on datasets with few or short documents.

Documents from the same domain usually discuss similar topics in a similar order. In this paper we present new ordering-based topic models that use generalised Mallows models to capture this regularity to constrain topic assignments. Specifically, these new models assume that there is a canonical topic ordering shared amongst documents from the same domain, and each document-specific topic ordering is allowed to vary from the canonical topic ordering. Instead of full orderings over a set of all possible topics covered by a domain, we make use of top-t orderings via a multistage ranking process. We show how to reformulate the new models so that a point-wise sampling algorithm from the Bayesian word segmentation literature can be used for posterior inference. Experimental results on several document collections with different properties show that our model performs much better than the other topic ordering-based models, and competitively with other state-of-the-art topic segmentation models.

K-nearest neighbor (KNN) rule is a simple and effective algorithm in pattern classification. In this article, we propose a local mean-based k-nearest centroid neighbor classifier that assigns to each query pattern a class label with nearest local centroid mean vector so as to improve the classification performance. The proposed scheme not only takes into account the proximity and spatial distribution of k neighbors, but also utilizes the local mean vector of k neighbors from each class in making classification decision. In the proposed classifier, a local mean vector of k nearest centroid neighbors from each class for a query pattern is well positioned to sufficiently capture the class distribution information. In order to investigate the classification behavior of the proposed classifier, we conduct extensive experiments on the real and synthetic data sets in terms of the classification error. Experimental results demonstrate that our proposed method performs significantly well, particularly in the small sample size cases, compared with the state-of-the-art KNN-based algorithms.

Understanding how topics within a document evolve over its structure is an interesting and important problem. In this paper, we address this problem by presenting a novel variant of Latent Dirichlet Allocation (LDA): Sequential LDA (SeqLDA). This variant directly considers the underlying sequential structure, i.e., a document consists of multiple segments (e.g., chapters, paragraphs), each of which is correlated to its previous and subsequent segments. In our model, a document and its segments are modelled as random mixtures of the same set of latent topics, each of which is a distribution over words; and the topic distribution of each segment depends on that of its previous segment, the one for first segment will depend on the document topic distribution. The progressive dependency is captured by using the nested two-parameter Poisson Dirichlet process (PDP). We develop an efficient collapsed Gibbs sampling algorithm to sample from the posterior of the PDP. Our experimental results on patent documents show that by taking into account the sequential structure within a document, our SeqLDA model has a higher fidelity over LDA in terms of perplexity (a standard measure of dictionary-based compressibility). The SeqLDA model also yields a nicer sequential topic structure than LDA, as we show in experiments on books such as Melville's "The Whale".

It is well known that either domain specific or domain independent knowledge has been adopted in Information retrieval (IR) to improve the retrieval performance. In this paper, we propose a novel IR model for digital forensics by using latent semantic indexing (LSI) and WordNet as an underlying reference ontology to retrieve suspicious emails according to the semantic meaning of an investigatorpsilas query. Our model incorporates corpus independent knowledge from WordNet and corpus dependent knowledge from LSI into query expansion and reduction; and LSI is also adopted to simulate human meaning based judgement of relatedness between investigatorpsilas queries and emails. We compare the performance of the resulting LSI And WordNet based Information retrieval system (LAWIRS) with other three systems we implement, i.e. the LSI system, the Lucene system and the Lucene system with query expansion. Experimental results on several email datasets demonstrate that for short Boolean queries, LAWIRS can successfully capture their meaning and yield substantial improvements in the overall retrieval performance.