Large Language Models and Knowledge Computing

Dear Colleagues,

Currently, the interdisciplinary research on large language models (LLMs) and knowledge computing is leading a new wave of development in the field of artificial intelligence. The core scientific challenge of this research direction lies in achieving deep synergy between structured knowledge (e.g., knowledge graphs and domain rules) and the implicit knowledge representation capabilities of large-scale pretrained models, as well as leveraging domain knowledge to overcome the cognitive boundaries and reasoning limitations of large models. This Special Issue focuses on (but is not limited to) the following topics:

Exploring the Knowledge Boundaries of Large Language Models

• Multi-dimensional evaluation of the inherent limitations of large language models in terms of knowledge coverage, comprehension depth, and reasoning capabilities across different domains and tasks.
• Investigating boundary-expanding methods based on data augmentation, knowledge injection, and architectural optimization.
• Developing frameworks to enhance model reliability, safety, and adaptability.

New Paradigms for Knowledge Mining Based on Large Language Models

• Novel methods for entity and relation extraction using large language models.
• Automated knowledge discovery and structuring techniques for massive unstructured data.
• Applications of large language models' semantic understanding and generation capabilities in identifying new concepts and uncovering latent relationships.

Construction of Interpretable and Verifiable Knowledge Systems to Support Intelligent Reasoning and Decision-Making

• Innovations in knowledge-enhanced model training.
• Optimization of pretraining objectives based on entity relations and logical constraints.
• Knowledge-driven solutions to mitigate hallucination issues in large language models.

Systematic Strategies to Improve the Trustworthiness of Generated Content

• Research on dynamic knowledge update mechanisms.
• Innovative solutions to address the static knowledge limitation of large models.
• Real-time retrieval-augmented techniques based on external knowledge bases.

Advanced Methods for Parametric Knowledge Rditing and Incremental Learning

• Breakthroughs in knowledge reasoning and interpretability.
• Hybrid reasoning frameworks combining symbolic reasoning and neural computation.
• Optimization methods for reasoning chains based on logical rules.

Knowledge Neuron Analysis and Interpretability Research on Model Decision Paths

• Active reasoning based on large language models.
• Proactive discovery and dynamic completion of missing knowledge in large language models.
• Non-chain decomposition and multi-path collaborative solving of complex problems using large language models.
• Exploration of autonomous cognitive evolution and reasoning enhancement mechanisms for large language models in open-world scenarios.

This series of research topics collectively forms a comprehensive framework for the integration and innovation of large language models and knowledge computing. It will drive the paradigm shift in AI from data-driven to knowledge-driven approaches, laying the theoretical foundation and providing technical support for building a new generation of trustworthy, reliable, and interpretable intelligent systems.

Prof. Dr. Ming Liu
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Applied Sciences is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• knowledge computing
• domain knowledge-enhanced large language models
• active reasoning based on large language models
• exploration of the knowledge boundaries of large language models