5 results about "Expert opinion" patented technology

The invention discloses a heuristic medical education large model fine tuning method and device based on clinical thinking. The method comprises the step of constructing a clinical thinking framework based on clinical teaching data and expert opinions. According to the heuristic medical education large model fine tuning method and device based on clinical thinking, a structured clinical teaching corpus is formed through a clinical thinking framework, a three-agent collaborative architecture of a teacher agent, an expert agent and a student agent is constructed based on the clinical teaching corpus, and a final data set is generated through the collaborative architecture. And training and fine tuning are carried out through the final data set to obtain the final medical education large model, so that the interactivity, adaptability and teaching guidance capability of the final medical education large model are remarkably improved. The final medical education large model can gradually guide students to conduct reasoning and exploration according to a clinical thinking framework, while the teaching logicality is kept, the independent thinking ability of users is effectively stimulated, moderate guidance and feedback are provided, and therefore the actual requirements of clinical teaching are met.

This invention discloses a method and apparatus for fine-tuning a large-scale medical education model based on clinical thinking, comprising constructing a clinical thinking framework based on clinical teaching data and expert opinions. This method and apparatus forms a structured clinical teaching corpus through the clinical thinking framework, and constructs a three-agent collaborative architecture—teacher agent, expert agent, and student agent—based on the corpus. The final dataset is generated through this collaborative architecture, and fine-tuned using the final dataset to obtain the final large-scale medical education model. This significantly improves the interactivity, adaptability, and teaching guidance capabilities of the final large-scale medical education model. The final model can gradually guide students to reason and explore based on the clinical thinking framework, effectively stimulating users' independent thinking ability while maintaining the logical flow of teaching, providing appropriate guidance and feedback, thus meeting the actual needs of clinical teaching.

The invention relates to a multivariate evidence correlation analysis method in the judicial field, and belongs to the technical field of intelligent processing of documents. The method comprises the following steps: loading judicial case file data and extracting an evidence content text; constructing an evidence rule template according to judicial domain knowledge and expert opinions; extracting evidence elements based on the rule template and constructing an evidence library; building a multivariate evidence information network, calculating the condition probability of association information among evidence elements based on a Bayesian network, and forming a multivariate evidence association analysis model; optimizing the multivariate evidence information network through a random strategy reinforcement learning algorithm; and calculating an information probability value of the multivariate evidence subnet, and selecting the most valuable multivariate evidence information network. According to the method, a multivariate evidence correlation analysis event of a judicial case is converted into an updating and optimizing task of a multivariate evidence information network, the complex correlation problem among multivariate evidences is simplified, a correlation model among the multivariate evidences is given, and certain auxiliary support is provided in a judicial decision support system.

The application provides a large language model legal field robustness evaluation method based on knowledge injection attack, wherein the method comprises the following steps: by searching for enhanced generation attack and similar charge attack, the large language model in the legal field is disturbed based on the large premise judgment of the input legal provisions; by vocabulary attack, element attack and narrative attack, the narrative accuracy of the large language model based on the input case facts is disturbed; by previous behavior attack and expert opinion attack, the final conclusion judgment of the large language model is disturbed to evaluate its anti-interference ability in legal reasoning. The application provides a logic attack framework based on Aristotle syllogism, attacks the semantic logic and professional logic reasoning of the problem, more widely covers the error types frequently occurring in real life, and can truly evaluate the robustness of the large model in the field.

The invention discloses a process evaluation method based on HDBSCAN clustering and a large language model, and the method comprises the steps: decomposing each expert questionnaire into the data of each evaluation index according to the characteristics of a project and the opinions of experts: the first questionnaire and the first evaluation index of the first questionnaire; the feature vector of each evaluation dimension of each evaluation index is constructed, the evaluation dimension is the first evaluation dimension of the first evaluation index of the first questionnaire, and the evaluation dimensions comprise the expert working age, the traditional process score and the intelligent construction process score. Complicated descriptive texts and a large number of expert questionnaire tables can be integrated to generate scientific and comprehensive process evaluation results and optimization suggestions, the logical reasoning ability of LLM on specific scene complex corpora is enhanced, and the LLM input length limitation is overcome.