Method for searching for related patent by using clinical trial design data based on large language model

A server for searching for a related patent by using clinical trial design data based on an LLM, comprising:

• a memory that stores a program configured to perform a method for searching for the related patent by using the clinical trial design data based on the LLM;
• and
• a processor that executes the program,
• wherein the method includes:
• (a) receiving a clinical trial-related keyword from a user device, searching for a plurality of pieces of clinical trial design data based on the clinical-trial-related keyword, and providing the plurality of pieces of clinical trial design data, wherein before performing process (a),
• (a-1) receiving a plurality of pieces of patent document data about a clinical trial from the user device;
• (a-2) extracting a query information string including a clinical trial title, a disease name, information about a clinical trial phase, clinical trial patient recruitment criteria, and clinical trial patient group design information, and a response information string including claims of a patent document, description of an invention of the patent document, patent bibliographic data, and a patent number from the plurality of pieces of patent document data, and generating a prompt command by using a pre-stored query text, the query information string, and the response information string;
• (a-3) matching the prompt command with the query information string and the response information string to create a training dataset and adjusting parameters of the LLM by training the LLM with the training dataset;
• and
• (a-4) performing validation by setting patent document data for validation as a separate validation dataset among the plurality of pieces of patent document data set by a user, training the LLM with the separate validation dataset, and adjusting the parameters of the LLM;
• (b) when one of the plurality of pieces of clinical trial design data is selectively input from the user device, providing the user device with detailed information about the selectively input clinical trial design data;
• (c) receiving a category of a main keyword from the user device and extracting a main keyword corresponding to the category from the detailed information about the clinical trial design data;
• (d) pre-processing the extracted main keyword, combining a plurality of pieces of pre-stored query text to generate a plurality of pieces of final query text, and inputting the final query text to the LLM to search for and collect a plurality of related patents;
• and
• (e) generating a report in which the plurality of related patents collected is output for a plurality of sections, respectively, and providing the report to the user device.