Knowledge recall and innovation hypothesis generation method based on solid waste cementing material large language model

Abstract

The invention relates to a knowledge recall and innovation hypothesis generation method based on a large language model of a solid waste cementing material, which realizes intelligent retrieval and knowledge integration of literatures, patents and experimental data in the field of the solid waste cementing material by constructing a special field corpus and finely adjusting the large language model, and breaks through the efficiency bottleneck of traditional manual retrieval. Further combining with a retrieval enhancement generation (RAG) technology and structured cue word design, the model can generate a material innovation hypothesis with a scientific basis based on multi-source knowledge association. Meanwhile, by integrating literature analysis, hypothesis generation, experimental design and a visual tool, an end-to-end research auxiliary process is constructed, and full-chain cooperation from theoretical exploration to experimental verification is supported. The knowledge acquisition and integration efficiency in the solid waste cementing material field is remarkably improved, and the problem of repeated research caused by literature dispersion is solved; a high-value research hypothesis is generated through data-driven intelligent reasoning, and the new material development period is effectively shortened.

Images