The static nature of knowledge within Large Language Models (LLMs) makes it difficult for them to adapt to evolving information, rendering knowledge editing a critical task. However, existing methods struggle with challenges of scalability and retrieval efficiency, particularly when handling complex, multi-hop questions that require multi-step reasoning. To address these challenges, this paper introduces ALEX (A Light Editing-knowledge Extractor), a lightweight knowledge editing framework. The core innovation of ALEX is its hierarchical memory architecture, which organizes knowledge updates (edits) into semantic clusters. This design fundamentally reduces retrieval complexity from a linear O(N) to a highly scalable O(K+N/C). Furthermore, the framework integrates an Inferential Query Synthesis (IQS) module to bridge the semantic gap between queries and facts , and a Dynamic Evidence Adjudication (DEA) engine that executes an efficient two-stage retrieval process. Experiments on the MQUAKE benchmark demonstrate that ALEX significantly improves both the accuracy of multi-hop answers (MultiHop-ACC) and the reliability of reasoning paths (HopWise-ACC). It also reduces the required search space by over 80% , presenting a promising path toward building scalable, efficient, and accurate knowledge editing systems.

Large language models (LLMs) have made remarkable progress in natural language processing tasks. However, efficiently integrating new knowledge and accurately answering multi-hop questions remain significant challenges. Existing methods often struggle with question semantic understanding and knowledge integration during reasoning. This paper proposes the ReAligner to address these issues. The ReAligner comprises three key components. The Multi-Representation Generator generates multiple semantically rich question representations by applying synonym replacement, syntactic variation, and semantic expansion, enhancing the model’s understanding of multi-hop questions. The Knowledge Relevance Filter filters out irrelevant facts from large fact sets, improving retrieval precision and computational efficiency. The Semantic Alignment Predictor determines the most relevant fact for each sub-question by leveraging the Semantic Sharpening Unit and a trained classifier. We evaluate the ReAligner model using four datasets and compare it with state-of-the-art methods. Experimental results show that the ReAligner model outperforms existing methods.

Abstract Enterprise credit risk prediction is to predict whether enterprises will default in the future, according to a variety of historical data by establishing a corresponding relationship between historical operating conditions and default status. To improve the accuracy of credit risk prediction of listed real estate enterprises and effectively reduce difficulty of government management, we propose an attention‐based CNN‐BiLSTM hybrid neural network enhanced with features of results of logistic regression, and constructs the credit risk prediction index system of listed real estate enterprises from five characteristic dimensions: profitability, debt‐paying ability, growth ability, operating ability and enterprise basic information. This study uses data from the 2017–2020 annual reports of listed real estate enterprises on China's Shanghai and Shenzhen stock exchanges. A five different verifications yields average sensitivity, specificity, and quality index of 99.28%, 94.57% and 97.15%, respectively. The results show that our approach achieves better experimental results than previous works, by comparing PSO‐SVM model, RS‐PSO‐SVR model and PSO‐BP model. We conclude that the Logistic‐CNN‐BiLSTM‐att model is more effective for the credit risk prediction of listed real estate enterprises.

In this paper we discuss a new approach for the detection of clustered microcalcifications (MCs) in mammograms. MCs are an important early sign of breast cancer in women. Their accurate detection is a key issue in computer aided detection scheme. To improve the performance of detection, we propose a Bagging and Boosting based twin support vector machine (BB-TWSVM) to detect MCs. The algorithm is composed of three modules: the image pro-processing, the feature extraction component and the BB-TWSVM module. The ground truth of MCs in mammograms is assumed to be known as a priori. First each MCs is preprocessed by using a simple artifact removal filter and a well designed high-pass filter. Then the combined image feature extractors are employed to extract 164 image features. In the combined image features space, the MCs detection procedure is formulated as a supervised learning and classification problem, and the trained BB-TWSVM is used as a classifier to make decision for the presence of MCs or not. The experimental results of this study indicate the potential of the approach for computer-aided detection of breast cancer.