[Purpose/Significance] Scholar profiles are widely applied in various precision services within the academic field, such as paper review, project evaluation, targeted recommendations, and talent recruitment. Topic tags play a critical role in identifying scholars in the same specialized subfield. This paper optimizes the method for calculating scholar topic tags to more accurately identify scholars within specific subdisciplines. [Method/Process] This paper proposed a theoretical framework: scholar profiles should comprehensively incorporate evaluation information such as scholars' research interests and academic capabilities, while also reflecting their research fields and academic standing within those fields. This paper integrated topic tags with evaluation data, using the latter as a factor for weighting topic tags. An algorithm was proposed for calculating high-scoring topic tags for scholars, incorporating three dimensions: the quantity of published papers, the author's position in the author list, and the citations of the papers. [Result/Conclusion] From the perspective of individual scholar profiling, high-score topic tags can delineate the main specialized research areas of a scholar. For the need to identify experts within a discipline, high-score topic tags can help pinpoint the most outstanding and core scholars in that field.
[Purpose/Significance] Based on the combination of traditional topic models and large language models, as well as RAG-based large language model, this study designs the topic modeling method, analyzes the diffusion of emerging topics and non-emerging topics, so as to provide a reference for the subsequent identification and diffusion of emerging topics. [Methods/Process] First of all, the topic mining of patents was carried out through the combination of traditional topic models and large language models, and the emerging topics were identified based on novelty, continuity, growth and influence. Secondly, the best topic modeling method was determined by using the large language model score fine-tuned by RAG. Finally, a citation network was constructed to calculate the diffusion speed, breadth, intensity, depth, and region, and analyze the temporal changes of the diffusion indicators of emerging topics and non-emerging topics. [Result/Conclusion] Taking nuclear fusion as the empirical field, the research finds that the topic modeling method based on BERTopic+Doubao has the best effect. At the same time, the diffusion of emerging themes mostly shows a fluctuating upward trend, while non-emerging themes show a steady upward trend first and then a slight downward trend. The method of combining the traditional topic model and the large language model proposed in this paper can effectively identify emerging topics. The RAG fine-tuning large language model optimizes the evaluation effect of the results, and enriches the perspective of topic diffusion measurement.
[Purpose/Significance] Against the background of the increasingly prominent reproducibility issues in digital humanities research, this paper systematically analyzes the challenges posed by data complexity, methodological heterogeneity, and the multimodal nature of research outputs to research verifiability, and explores the critical role of academic journals in institutional regulation and technical governance. [Method/Process] From the three dimensions of data, methods, and research outputs, the paper examined the dilemmas of reproducibility and, on this basis, introduced an analytical framework of “boundary-keeping, boundary-breaking, and their dynamic balance”. By integrating data publishing, platform-based collaboration, intelligent empowerment, and human-machine collaboration mechanisms, it constructed practical pathways through which journals can enhance reproducibility. [Result/Conclusion] It is argued that journals should achieve transparent and traceable management of the entire research process through standardized data collection and metadata specifications, centralized platform management, intelligent quality control, and human–machine collaborative peer review. The study concludes that academic journals need to maintain a dynamic balance between safeguarding academic norms and promoting publishing innovation, gradually forming an integrated “data–methods–publishing” ecosystem, so as to enhance the credibility, reproducibility, and academic impact of digital humanities research, and to provide institutional support for building an open, transparent, and sustainable scholarly publishing system.
[Purpose/Significance] People can quickly obtain the necessary knowledge from a vast amount of information, relying on knowledge to represent complex structured and semantic modeling of information. This article describes knowledge from multiple perspectives, constructs a multi-perspective knowledge representation framework model, and explores the innovation of knowledge organization theories and methods, with the aim of further promoting the strategic development of artificial intelligence. [Method/Process] Knowledge representation perspectives were classified into perspectives such as knowledge object, knowledge subject, and knowledge situation based on different perspectives. The theoretical basis of the methods under these knowledge representation perspectives was described. The framework and model of multi-perspective knowledge representation were constructed, the formal representation of knowledge order was realized, and a brand-new operating mechanism of multi-perspective knowledge representation was proposed. [Result/Conclusion] To maximize the effectiveness of knowledge representation in a digital and intelligent environment, multi-perspective knowledge representation can be adjusted in aspects such as establishing a flexible and multi-dimensional holographic knowledge order, integrating multi-modal and contextualized expressions, and dynamically evolving for user problems, so as to better serve specific practical fields.
[Purpose/Significance] In response to the practical demand for high-quality data support and trustworthy operation of domain large models in specialized and high-risk application scenarios, this study systematically examines the basic connotation, key characteristics, and systematic construction path of the domain large model data foundation. It aims to provide a theoretical framework and methodological support for the practical deployment of trustworthy artificial intelligence, and to offer systematic references for the standardized construction and governance practice of the domain large model data foundation. [Method/Process] Based on a systematic review of the relevant literature, this study clarified the conceptual connotation, major characteristics, and constituent elements of the domain large model data foundation. It introduced the theory of layered modular architecture to construct a layered framework for the domain large model data foundation, mainly covering the data resource layer, knowledge organization layer, knowledge storage and semantic indexing layer, model layer, algorithm layer, application layer, and governance and operational support layer, so as to characterize the overall mechanism of the coordinated operation of data, knowledge, and models. Finally, a desk research method based on public evidence and a cross-case comparison approach were employed to demonstrate the contextual applicability of the framework. [Result/Conclusion] Cross-domain comparisons show that the proposed framework can explain the common mechanisms underlying external knowledge memory access, evidence-constrained generation, and compliance-audit output chains in reference systems across different domains. On this basis, the study further distills reusable methodological elements such as data contracts, index configuration, and evaluation baselines, thereby validating the applicability and transferability of the framework across domains.
