Embedded Files
 
施育廷(2026)。基於對話式介面之多源資訊整合與自動化內容生成之系統與方法… | Andrew Shih施育廷(2026)。基於對話式介面之多源資訊整合與自動化內容生成之系統與方法 (I918530)。中華民國經濟部智慧財產局。 本發明提供一種基於對話式介面之多源資訊整合與自動化內容生成之系統與方法。該方法包含:由一後端邏輯伺服器接收一使用者透過一通訊軟體發出之一第一指令;該後端邏輯伺服器傳送該第一指令至一大型語言模型以進行意圖分析,並獲得一包含一任務意圖及至少一參數之結構化指令;該後端邏輯伺服器根據該任務意圖,調用至少一內部資訊源及一外部資訊源,以分別獲取內部數據與外部數據;以及,該後端邏輯伺服器將該內部數據與該外部數據整合為一綜合性上下文,並傳送至該大型語言模型以生成一最終產出,再將該最終產出回傳至該使用者。藉此,本發明解決了現有技術中資訊孤島與手動整合耗時之問題,顯著提升知識工作者的研究與內容創作效率。 This invention provides a system and method for multi-source information integration and automated content generation based on a conversational interface. The method comprises: receiving a first command from a user via a communication software by a backend logic server; sending the first command to a large language model for intent analysis to obtain a structured command including a task intent and at least one parameter; invoking at least one internal information source and one external information source by the backend logic server according to the task intent to acquire internal data and external data, respectively; and integrating the internal data and the external data into a comprehensive context, sending it to the large language model to generate a final output, and returning the final output to the user. Thereby, the present invention solves the problems of information silos and time-consuming manual integration in the prior art, significantly improving the research and content creation efficiency for knowledge workers.
Google Sites
Report abuse