Optimizing the Path of AIGC Creative Content Generation Based on Large Language Models and Knowledge Graphs

Yushu Cao
Proceedings of 2025 2nd International Conference on Machine Learning and Intelligent Computing, PMLR 278:550-559, 2025.

Abstract

With the rapid development of generative artificial intelligence (AIGC) technologies, creative content generation using large language models (LLMs) and knowledge graphs (KGs) has become a key area of research. However, current methods often fail to fully harness the semantic enhancement capabilities of knowledge graphs in the content generation process. To address this gap, this paper proposes an innovative creative content generation path optimization model, KG-GPT-opt, which integrates large language models with knowledge graphs. Experimental results demonstrate that the KG-GPT-opt model outperforms traditional baseline models across several standard evaluation metrics, achieving improvements of 6.4%, 7.5%, and 4.8% in BLEU, ROUGE-L, and METEOR, respectively. Furthermore, the model receives high expert ratings of 8.6 and 8.8 for creativity and coherence, surpassing other generation models. This study offers a novel approach for the AIGC field, advancing the integration of large language models and knowledge graphs, and broadening the potential applications of intelligent content generation in areas such as cultural creativity and smart marketing.

Cite this Paper

BibTeX
@InProceedings{pmlr-v278-cao25a, title = {Optimizing the Path of AIGC Creative Content Generation Based on Large Language Models and Knowledge Graphs}, author = {Cao, Yushu}, booktitle = {Proceedings of 2025 2nd International Conference on Machine Learning and Intelligent Computing}, pages = {550--559}, year = {2025}, editor = {Zeng, Nianyin and Pachori, Ram Bilas and Wang, Dongshu}, volume = {278}, series = {Proceedings of Machine Learning Research}, month = {25--27 Apr}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v278/main/assets/cao25a/cao25a.pdf}, url = {https://proceedings.mlr.press/v278/cao25a.html}, abstract = {With the rapid development of generative artificial intelligence (AIGC) technologies, creative content generation using large language models (LLMs) and knowledge graphs (KGs) has become a key area of research. However, current methods often fail to fully harness the semantic enhancement capabilities of knowledge graphs in the content generation process. To address this gap, this paper proposes an innovative creative content generation path optimization model, KG-GPT-opt, which integrates large language models with knowledge graphs. Experimental results demonstrate that the KG-GPT-opt model outperforms traditional baseline models across several standard evaluation metrics, achieving improvements of 6.4%, 7.5%, and 4.8% in BLEU, ROUGE-L, and METEOR, respectively. Furthermore, the model receives high expert ratings of 8.6 and 8.8 for creativity and coherence, surpassing other generation models. This study offers a novel approach for the AIGC field, advancing the integration of large language models and knowledge graphs, and broadening the potential applications of intelligent content generation in areas such as cultural creativity and smart marketing.} }
Endnote
%0 Conference Paper %T Optimizing the Path of AIGC Creative Content Generation Based on Large Language Models and Knowledge Graphs %A Yushu Cao %B Proceedings of 2025 2nd International Conference on Machine Learning and Intelligent Computing %C Proceedings of Machine Learning Research %D 2025 %E Nianyin Zeng %E Ram Bilas Pachori %E Dongshu Wang %F pmlr-v278-cao25a %I PMLR %P 550--559 %U https://proceedings.mlr.press/v278/cao25a.html %V 278 %X With the rapid development of generative artificial intelligence (AIGC) technologies, creative content generation using large language models (LLMs) and knowledge graphs (KGs) has become a key area of research. However, current methods often fail to fully harness the semantic enhancement capabilities of knowledge graphs in the content generation process. To address this gap, this paper proposes an innovative creative content generation path optimization model, KG-GPT-opt, which integrates large language models with knowledge graphs. Experimental results demonstrate that the KG-GPT-opt model outperforms traditional baseline models across several standard evaluation metrics, achieving improvements of 6.4%, 7.5%, and 4.8% in BLEU, ROUGE-L, and METEOR, respectively. Furthermore, the model receives high expert ratings of 8.6 and 8.8 for creativity and coherence, surpassing other generation models. This study offers a novel approach for the AIGC field, advancing the integration of large language models and knowledge graphs, and broadening the potential applications of intelligent content generation in areas such as cultural creativity and smart marketing.
APA
Cao, Y.. (2025). Optimizing the Path of AIGC Creative Content Generation Based on Large Language Models and Knowledge Graphs. Proceedings of 2025 2nd International Conference on Machine Learning and Intelligent Computing, in Proceedings of Machine Learning Research 278:550-559 Available from https://proceedings.mlr.press/v278/cao25a.html.

Related Material