While Large Language Models (LLMs) excel at many reasoning tasks, their native inability to produce calibrated, multi-class probability distributions limits their use in high-stakes Web applications like content moderation and fraud detection. Existing methods to elicit probabilities from LLMs either sacrifice their crucial Chain-of-Thought (CoT) reasoning capabilities or suffer from poor calibration. To address this, we introduce a new paradigm, Verbalized Probability Distribution, and a novel training framework, RLVP (Reinforcement Learning with Verbalized Probabilities). RLVP fine-tunes an LLM to generate both an interpretable CoT and a complete, verbalized probability distribution. We overcome the ''insufficient reward granularity'' problem in standard Reinforcement Learning (RL) for classification by using soft probabilities from expert tabular models as a dense reward curriculum. Through large-scale joint training on 169 tabular tasks, we demonstrate that a single RLVP-trained model can surpass a strong, task-specific XGBoost baseline on up to 55% of tasks. More importantly, the trained model achieves state-of-the-art few-shot performance on unseen, heterogeneous Web benchmarks that mix structured data with free text, achieving performance comparable to or superior than expert models trained on the same limited data. This showcases a strong capability for generalization and knowledge transfer to complex Web data. Our work presents a viable path toward building general-purpose, probabilistically-sound, and interpretable foundation models for the Web.

Log-based anomaly detection plays a crucial role in ensuring the reliability of systems. While deep learning-based small detection models (SDMs) are efficient, the large language models (LLMs) are accurate and capable of providing explanations. Intuitively, a compelling question arises: Can we seamlessly combine the advantages of both approaches? In this work, we delve into this underexplored research direction and propose CoLA, a novel collaborative log anomaly detection framework. During collaborative inference, an SDM serves as a filter to select potentially anomalous instances, while a downstream LLM acts as an expert to detect anomalies, offer explanations, and refine the SDM. Extensive experiments on three large real-world datasets demonstrate that CoLA significantly outperforms state-of-the-art methods in terms of effectiveness, efficiency, and explainability, while also greatly reducing labor costs.

Multi-label text classification (MLTC) allows a given text to be associated with multiple labels, which well suits many real-world data mining scenarios. However, the annotation effort of MLTC is inevitably expensive and time-consuming. Although multi-label active learning provides a cost-effective solution, it still faces two major challenges: (i) constructing decent feature space to distinguish the confusing semantics of different labels; (ii) defining proper sampling criteria to measure a sample's joint effect over the entire label space. To bridge these gaps, we propose a Contrastive Multi-label Active Learning framework (CoMAL) that gives an effective data acquisition strategy. Specifically, a contrastive decoupling mechanism is introduced to fully release the semantic information of multiple labels into the latent space. Then, we devise a hybrid criterion that balances two data value measures: (i) similarity-enhanced label cardinality inconsistency reflects the uncertainty of data predictions. (ii) positive feature diversity evaluates the positive-propensity semantic diversity to handle the label sparsity. Extensive experiments demonstrate that our CoMAL outperforms the current state-of-the-art multi-label active learning approaches. Code for CoMAL is available at https://github.com/chengzju/CoMAL.

Tabular data pervades the landscape of the World Wide Web, playing a foundational role in the digital architecture that underpins online information. Given the recent influence of large-scale pretrained models like ChatGPT and SAM across various domains, exploring the application of pretraining techniques for mining tabular data on the web has emerged as a highly promising research direction. Indeed, there have been some recent works around this topic where most (if not all) of them are limited in the scope of a fixed-schema/single table. Due to the scale of the dataset and the parameter size of the prior models, we believe that we have not reached the ''BERT moment'' for the ubiquitous tabular data. The development on this line significantly lags behind the counterpart research domains such as natural language processing. In this work, we first identify the crucial challenges behind tabular data pretraining, particularly overcoming the cross-table hurdle. As a pioneering endeavor, this work mainly (i)-contributes a high-quality real-world tabular dataset, (ii)-proposes an innovative, generic, and efficient cross-table pretraining framework, dubbed as CM2, where the core to it comprises a semantic-aware tabular neural network that uniformly encodes heterogeneous tables without much restriction and (iii)-introduces a novel pretraining objective --- prompt Masked Table Modeling (pMTM) --- inspired by NLP but intricately tailored to scalable pretraining on tables. Our extensive experiments demonstrate CM2's state-of-the-art performance and validate that cross-table pretraining can enhance various downstream tasks.

In the era of big data, data annotation is integral to numerous applications. However, it is widely acknowledged as a laborious and time-consuming process, significantly impeding the scalability and efficiency of data-driven applications. To reduce the human cost, we demonstrate CORAL, a collaborative automatic labeling system driven by large language models (LLMs), which achieves high-quality annotation with the least human effort. Firstly, CORAL employs LLM to automatically annotate vast datasets, generating coarse-grained labels. Subsequently, a weakly-supervised learning module trains small language models (SLMs) using noisy label learning techniques to distill accurate labels from LLM's annotations. It also allows statistical analysis of model outcomes to identify potentially erroneous labels, reducing the human cost of error detection. Furthermore, CORAL supports iterative refinement by LLMs and SLMs using manually corrected labels, thereby ensuring continual enhancement in annotation quality and model performance. A visual interface enables annotation process monitoring and result analysis.

Multiple-instance learning (MIL) is a significant weakly supervised learning problem, where the training data consist of bags containing multiple instances and bag-level labels. Most previous MIL research required fully labeled bags. However, collecting such data is challenging due to the labeling costs or privacy concerns. Fortunately, we can easily collect pairwise comparison information, indicating one bag is more likely to be positive than the other. Therefore, we investigate a novel MIL problem about learning a bag-level binary classifier only from pairwise comparison bags. To solve this problem, we display the data generation process and provide a baseline method to train an instance-level classifier based on unlabeled-unlabeled learning. To achieve better performance, we propose a convex formulation to train a bag-level classifier and give a generalization error bound. Comprehensive experiments show that both the baseline method and the convex formulation achieve satisfactory performance, while the convex formulation performs better. 1

A formidable challenge in the multi-label text classification (MLTC) context is that the labels often exhibit a long-tailed distribution, which typically prevents deep MLTC models from obtaining satisfactory performance. To alleviate this problem, most existing solutions attempt to improve tail performance by means of sampling or introducing extra knowledge. Data-rich labels, though more trustworthy, have not received the attention they deserve. In this work, we propose a multiple-stage training framework to exploit both model- and feature-level knowledge from the head labels, to improve both the representation and generalization ability of MLTC models. Moreover, we theoretically prove the superiority of our framework design over other alternatives. Comprehensive experiments on widely used MLTC datasets clearly demonstrate that the proposed framework achieves highly superior results to state-of-the-art methods, highlighting the value of head labels in MLTC.

Feature engineering often plays a crucial role in building mining systems for tabular data, which traditionally requires experienced human experts to perform. Thanks to the rapid advances in reinforcement learning, it has offered an automated alternative, i.e. automated feature engineering (AutoFE). In this work, through scrutiny of the prior AutoFE methods, we characterize several research challenges that remained in this regime, concerning system-wide efficiency, efficacy, and practicality toward production. We then propose Catch, a full-fledged new AutoFE framework that comprehensively addresses the aforementioned challenges. The core to Catch composes a hierarchical-policy reinforcement learning scheme that manifests a collaborative feature engineering exploration and exploitation grounded on the granularity of the whole feature set. At a higher level of the hierarchy, a decision-making module controls the post-processing of the attained feature engineering transformation. We extensively experiment with Catch on 26 academic standardized tabular datasets and 9 industrialized real-world datasets. Measured by numerous metrics and analyses, Catch establishes a new state-of-the-art, from perspectives performance, latency as well as its practicality towards production. Source code1 can be found at https://github.com/1171000709/Catch.

With the increase of dirty data, data cleaning turns into a crux of data analysis. The accuracy limitation of the existing integrity constraints-based cleaning approaches results from insufficient rules. In this paper, we present a novel hybrid data cleaning framework on top of Markov logic networks (MLNs), termed as${\sf MLNClean}$, which is capable of learning instantiated rules to supplement the insufficient integrity constraints.${\sf MLNClean}$consists of two steps, i.e.,pre-processingandtwo-stage data cleaning. In the pre-processing step,${\sf MLNClean}$first infers a set of probable instantiated rules according to MLNs and then builds a two-layer MLN index structure to generate multiple data versions and facilitate the cleaning process. In the two-stage data cleaning step,${\sf MLNClean}$first presents a concept ofreliability scoreto clean errors within each data version separately, and afterward eliminates the conflict values among different data version using a novel concept offusion score. Considerable experimental results on both real and synthetic scenarios demonstrate the effectiveness of${\sf MLNClean}$in practice.

With the growth of dirty data, data cleaning turns into a crux of data analysis. In this paper, we propose a novel hybrid data cleaning framework, termed as MLNClean, which is capable of learning instantiated rules to supplement the insufficient integrity constraints. MLNClean consists of two steps, i.e., pre-processing and two-stage data cleaning. In the pre-processing step, MLNClean first infers a set of probable instantiated rules according to Markov logic network (MLN) and then builds a two-layer MLN index to generate multiple data versions and facilitate the cleaning process. In the two-stage data cleaning step, MLNClean first presents a concept of reliability score to clean errors within each data version separately, and then, it eliminates the conflict values among different data versions using a novel concept of fusion score. Considerable experimental results on both real and synthetic scenarios demonstrate the effectiveness of MLNClean.