Extracting structured procedural knowledge from unstructured business documents is a critical yet unresolved bottleneck in process automation. While prior work has focused on extracting linear action flows from instructional texts (e.g., recipes), it has insufficiently addressed the complex logical structures—such as conditional branching and parallel execution—that are pervasive in real-world regulatory and administrative documents. Furthermore, existing benchmarks are limited by simplistic schemas and shallow logical dependencies, restricting progress toward logic-aware large language models (LLMs). To bridge this “Logic Gap”, we introduce BREX, a carefully curated benchmark comprising 409 real-world business documents and 2,855 expert-annotated rules. Unlike prior datasets centered on narrow service scenarios, BREX spans over 30 vertical domains, covering scientific, industrial, administrative, and financial regulations.We further propose ExIde, a structure-aware reasoning framework that investigates five distinct prompting strategies, ranging from implicit semantic alignment to executable grounding via pseudo-code generation, enabling explicit modeling of rule dependencies and providing an out-of-the-box framework for different business customers without finetuning their own LLMs. We benchmark ExIde using 13 state-of-the-art LLMs. Our extensive evaluation reveals that: (1) Executable grounding serves as a superior inductive bias, significantly outperforming standard prompts in rule extraction; and (2) Reasoning-optimized models demonstrate a distinct advantage in tracing long-range dependencies and non-linear rule dependencies compared to standard instruction-tuned models.

Existing Grammatical Error Correction (GEC) methods often overlook the assessment of sentence-level syntax and semantics in the corrected sentence. This oversight results in final corrections that may not be acceptable in the context of the original sentence. In this paper, to improve the performance of Grammatical Error Correction methods, we propose the post-processing task of Correction Acceptability Discrimination (CAD) which aims to remove invalid corrections by comparing the source sentence and its corrected version from the perspective of “sentence-level correctness”. To solve the CAD task, we propose a pipeline method where the acceptability of each possible correction combination based on the predicted corrections for a source sentence will be judged by a discriminator. Within the discriminator, we design a symmetrical comparison operator to overcome the conflicting results that might be caused by the sentence concatenation order. Experiments show that our method can averagely improve F_0.5 score by 1.01% over 13 GEC systems in the BEA-2019 test set.

Named Entity Recognition(NER), as a crucial subtask in natural language processing(NLP), is limited to a few labeled samples(a.k.a. few-shot). Metric-based meta-learning methods aim to learn the semantic space and assign the entity to its nearest label based on the similarity of their representations. However, these methods have trouble with semantic space learning and result in suboptimal performance. Specifically, the label name or its description is widely used for label semantic representation learning, but the label information extracted from the existing label description is limited. In addition, these methods focus on reducing the distance between the entity and the corresponding label, which may also reduce the distance between the labels and thus cause misclassification. In this paper, we propose a few-shot NER method that harnesses the power of Knowledge Graph and Contrastive Learning to improve the prototypical semantic space learning. First, KCL leverages knowledge graphs to provide rich and structured label information for label semantic representation learning. Then, KCL introduces the idea of contrastive learning to learn the label semantic representation. The label semantic representation is used to help distance the label clusters in the prototypical semantic space to reduce misclassification. Extensive experiments show that KCL achieves significant improvement over the state-of-the-art methods.

In this paper, we introduce Holistic Semantic Embedding and Global Contrast (HS-GC), an end-to-end approach to learn the instance- and cluster-level representation. Specifically, for instance-level representation learning, we introduce a new loss function that exploits different layers of semantic information in a deep neural network to provide a more holistic semantic text representation. Contrastive learning is applied to these representations to improve the model’s ability to represent text instances. Additionally, for cluster-level representation learning we propose two strategies that utilize global update to construct cluster centers from a global view. The extensive experimental evaluation on five text datasets shows that our method outperforms the state-of-the-art model. Particularly on the SearchSnippets dataset, our method leads by 4.4% in normalized mutual information against the latest comparison method. On the StackOverflow and TREC datasets, our method improves the clustering accuracy of 5.9% and 3.2%, respectively.

Named Entity Recognition (NER), as a crucial subtask in natural language processing (NLP), suffers from limited labeled samples (a.k.a. few-shot). Meta-learning methods are widely used for few-shot NER, but these existing methods overlook the importance of label dependency for NER, resulting in suboptimal performance. However, applying meta-learning methods to label dependency learning faces a special challenge, that is, due to the discrepancy of label sets in different domains, the label dependencies can not be transferred across domains. In this paper, we propose the Task-adaptive Label Dependency Transfer (TLDT) method to make label dependency transferable and effectively adapt to new tasks by a few samples. TLDT improves the existing optimization-based meta-learning methods by learning general initialization and individual parameter update rule for label dependency. Extensive experiments show that TLDT achieves significant improvement over the state-of-the-art methods.