Page 130 - AI for Good Innovate for Impact
P. 130

AI for Good Innovate for Impact



                      Partners:

                      Yunnan Baiyao Group: Focus on various scenarios including traditional Chinese medicine
                      (TCM) R&D, decision support, knowledge popularization, and digital marketing.

                      Huawei: Provides unified digital infrastructure, and Pangu Large Language Model solution.


                      2�2     Benefit of use case

                      •    Addressing patients' conditions, leveraging learning outcomes from over 10 million
                           pieces of TCM literature and millions of structured prescriptions, enhance the judgment
                           capabilities in complex scenarios such as TCM disease reasoning, providing precise drug
                           recommendations and treatment plans. Answer public questions about TCM, improve
                           public health literacy, and help people better prevent diseases and manage their health.
                      •    Through the multi-modal capabilities of the Leigong large model, quickly identify the
                           growth status of Chinese medicinal materials, prevent the impact of pests and diseases,
                           and provide professional solutions; farmers can follow the intelligent planting GAP (Good
                           Agricultural Practice) guidance at any time. By solving challenges such as data quality and
                           incremental training forgetting in TCM R&D, accelerate the development process of new
                           drugs and therapies, and promote the innovative development of TCM.


                      2�3     Future Work

                      The future work of TCM Industry Leigong Large Model includes:

                      1. Innovative application scenarios: Improve model development in three types of scenarios:
                      intelligent planting, intelligent diagnosis, and intelligent supply chain, such as intelligent
                      planting scenario tools, intelligent diagnosis of Chinese medicinal materials growth, and
                      intelligent consultation for TCM.

                      2. High-quality dataset: Build high-quality training and evaluation datasets for the scenario-
                      based solutions of TCM industry and medical health services. Regularly update the datasets to
                      adapt to the evolving intelligent technology and market demand changes in TCM.
                      3. Continuous operation: Establish a trusted data space for sustainable operations, operational
                      organization structure, and resource allocation mechanism.

                      4. Industry standards: Develop standards covering the application specifications, data models,
                      data dictionaries, data security, and privacy protection guidelines for AI models in the cultivation
                      and quality of TCM, medical health services, and health knowledge popularization, laying
                      a foundation for the healthy and orderly development of the industry. Data models and
                      data dictionaries standards will be developed in ITU-T SG20, while TCM industry AI model
                      application specifications will be developed in ITU-T SG21.


                      3      Use Case Requirements

                      REQ-01: Constructing a Traditional Chinese Medicine (TCM) vector database based on TCM
                      literature, prescriptions, and other TCM-related data. The TCM vector database is being
                      developed in a phased and compliant manner. Currently, it is in a restricted access mode,
                      open only to partner customers. It will be fully opened in the future.








                  94
   125   126   127   128   129   130   131   132   133   134   135