In knowledge graph construction, traditional NLP models suffer from poor cross-domain adaptability, while large language models (LLMs) face issues of hallucinations and data shift in information extraction. To address these challenges, we propose a collaborative framework for LLM-based triple extraction that integrates dynamic prompt generation and BERT assistance, effectively combining the entity recognition strengths of traditional models with the end-to-end reasoning capabilities of LLMs. On one hand, the framework uses the K-nearest neighbor (KNN) algorithm to vectorize text entity types. It matches highly relevant examples and generates dynamic prompts with progressive reasoning chain-of-thought (CoT) templates, enhancing LLMs' ability to adapt to task features and perform logical reasoning. On the other hand, it fine-tunes pre-trained BERT models to achieve accurate recognition and type annotation of text entities, alleviating LLM hallucinations by enhancing the structured representation of input text. Experimental validation on 3 cross-domain and cross-lingual datasets shows that when few-shot learning is combined with the CoT strategy, the triple extraction F1 score of Qwen2.5-14B increases from 0.1718 to 0.5856 (a 240.9 % relative improvement), demonstrating a performance breakthrough from strategy integration. After introducing BERT assistance, the F1 score of DeepSeek-V3 reaches 0.7426, and Qwen2.5-14B achieves a 277.3 % relative improvement compared to the zero-shot baseline, which verifies the effectiveness of cross-model collaboration. Compared with the traditional NLP model, it significantly enhances LLMs' knowledge transfer capability and domain adaptability. It provides an innovative paradigm for intelligent knowledge graph construction through the collaboration of traditional NLP techniques and prompt engineering, effectively compensating for the inherent limitations of LLMs.

For multiuser search scenarios over encrypted cloud data, the multiuser searchable encryption (MUSE) technology is a viable solution for incorporating privacy protection and favorable operability functions together. However, current MUSE technologies suffer from three practical limitations: 1) a secure channel between data owner and users will greatly increase the burden of owner; 2) a trusted third party (TTP) is required in some solutions; and 3) some solutions cannot resist keyword guessing attack (KGA), which means that the information of search trapdoor is easily leaked. In order to solve the above three problems, our first practical solution (DuMSE) without secure channel and without TTP is proposed in this article, which also can resist KGA. With inverted indexes, the search time complexity of DuMSE can be reduced to$r$, where$r$is the number of corresponding data. To support dynamically updating data, our second dynamic solution based on DuMSE is also proposed in this article, named as dynamic DuMSE (dDuMSE). The construction of dDuMSE includes not only the above features of DuMSE but also a new feature. Our analysis and experimental results show that the proposed schemes are secure and practical.

With the outbreak of e-mail message leakage events, such as the Hillary Clinton’s Email Controversy, privacy and security of sensitive e-mail information have become users’ primary concern. Encrypted email seems to be a viable solution for providing security, but it will greatly limit their operations. Public encryption with keyword search (PEKS) scheme is a popular technology to incorporate security protection and favorable operability functions together, which can play an important role in searching over encrypted email in a cloud server. In this paper, we propose a practical PEKS scheme named as public-key multi-keyword searchable encryption with hidden structures (PMSEHS). It could enable e-mail receivers to do the multi-keyword and boolean search in the large encrypted email database as fast as possible, without revealing more information to the cloud server. We also give comparative experiments, which demonstrate that our scheme has a higher efficiency in multi-keyword search for encrypted emails.