A knowledge-based question answering system based on car manuals

基于汽车手册的知识问答系统

• 批准号: 486046-2015
• 负责人: An, Aijun
• 金额: $ 1.82万
• 依托单位: York University
• 依托单位国家: 加拿大
• 项目类别: Engage Grants Program
• 财政年份: 2015
• 资助国家: 加拿大
• 起止时间: 2015-01-01 至 2016-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=585357
• 关键词: knowledge; based; question; answering; system

iNAGO is an international company dedicated to making computers more natural by providing a conversational assistant platform called netpeople. The netpeople platform combines innovative natural language conversational technology with tools and pre-set applications that smart device makers, such as Toyota, NEC, Clarion, and others, use to enable their own unique smart assistant service. One of iNAGO's key applications is a voice based question and answer (QA) system that will converse with users to obtain the most appropriate answers to their question. Companies such as Brother Industries, Toyota and others are currently using this system. iNAGO is looking to work with York to drastically improve this system and meet the needs of its customers. The York and iNAGO research team will employ a stepwise approach to build and test an in-car interactive QA system. This system works as the driver's navigator and digital concierge, enabling them to safely manage information and drive effectively at the same time. One of the major challenges in developing such an interactive QA system is the design and generation of a knowledge base for question answering. The current knowledge base used in iNAGO QA systems was mainly manually created, which was very time-consuming and not easy to maintain. In this Engage project, we will use machine learning techniques (such as topic modeling) to automatically create a knowledge base for an in-car QA system based on a car manual in order to answer questions that the user may have regarding the usage, operation and maintenance of the car. We will test our knowledge base by processing users' questions and providing answers to the users.

iNAGO 是一家国际公司，致力于通过提供名为 netpeople 的对话助理平台使计算机变得更加自然。 netpeople 平台将创新的自然语言对话技术与工具和预设应用程序相结合，丰田、NEC、Clarion 等智能设备制造商使用这些工具和预设应用程序来实现自己独特的智能助理服务。 iNAGO 的关键应用之一是基于语音的问答 (QA) 系统，该系统将与用户对话以获得最合适的问题答案。兄弟工业、丰田等公司目前正在使用该系统。 iNAGO 希望与约克合作，大幅改进该系统并满足客户的需求。 约克和 iNAGO 研究团队将采用逐步方法来构建和测试车内交互式 QA 系统。该系统充当驾驶员的导航器和数字礼宾员，使他们能够安全管理信息并同时有效驾驶。开发这种交互式问答系统的主要挑战之一是设计和生成用于问答的知识库。目前iNAGO QA系统使用的知识库主要是手动创建的，非常耗时且不易维护。 在这个Engage项目中，我们将利用机器学习技术（例如主题建模），根据汽车手册自动创建车内QA系统的知识库，以回答用户在汽车的使用、操作和维护方面可能存在的问题。我们将通过处理用户的问题并向用户提供答案来测试我们的知识库。