Core Insights - Shandong Province Information Industry Association announced the results of the 2025 "Shandong Province Information Industry Association Science and Technology Innovation Award," with Inspur Technology winning the first prize in the innovation invention category for its project on a low-rank compression-based large language model acceleration method [1][2] - The second prize in the technology achievement transformation category was awarded to Inspur Technology and Shandong University for their project on semantic-driven data governance intelligent agent key technologies and applications [1][4] Group 1: Innovation Invention - The project "A Low-Rank Compression-Based Large Language Model Acceleration Method and Implementation Device" addresses three core challenges in the deployment of large AI models: improving the quality of industry knowledge corpus, enhancing training and inference efficiency, and optimizing complex task reasoning [4][5] - The innovation includes the development of high-quality governance technology for industry corpus, significantly improving cross-domain data recognition accuracy [4] - The acceleration technology for large model training and inference doubles the efficiency of the entire training, inference, and deployment process, reducing deployment time from hours to minutes [4][5] Group 2: Technology Achievement Transformation - The project "Semantic-Driven Data Governance Intelligent Agent Key Technologies and Applications" focuses on challenges in data governance during digital transformation, such as the diversity of the industry, depth of knowledge, and quality issues [5] - The DataAgent platform integrates large models, rule engines, intelligent agents, and knowledge graph technologies to provide intelligent, automated, and secure data governance solutions [5] - A unified semantic framework is established to integrate multi-modal industry knowledge, creating a "semantic-logic-physical" three-layer mapping system to eliminate cognitive barriers [5]