When Search Goes Cognitive

当搜索变得认知化时

• 批准号: RGPIN-2019-06686
• 负责人: Hu, Qinmin
• 金额: $ 2.4万
• 依托单位: Ryerson University
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2022
• 资助国家: 加拿大
• 起止时间: 2022-01-01 至 2023-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=749829
• 关键词: When; Search; Goes; Cognitive

A goal of this work is to combine search, knowledge graph, question answering and recommendation to cognitive search on heterogeneous corpus. The proposed cognitive search architecture is presented to combine powerful indexing technology with advanced natural language processing capabilities and machine learning algorithms to build an increasingly deep corpus of knowledge from which to feed relevant information to users.  Three themes for this research work are proposed, namely data-driven knowledge graph, question answering system and recommendation respectively. 15 students, including both graduate and undergraduate students, will be trained. One Ph.D. student and one master's student will work on each project, and the undergraduate students mainly focus on search engines as their fourth-year projects or theses. All methods will be evaluated on real data sets such as Clueweb12-B13, Amazon, Microblog, Yelp, SemEval etc. Prototypes will be developed at every stage of the there themes. The results of this proposed research will be useful for Canadian, international businesses and government institutions. The proposed frameworks can be used by finance industry (e.g., TD Bank and stock market), healthcare (e.g., ICES), governmental institutions (e.g., Statistics Canada), and technological companies (e.g., Amazon and Google). Our program will train students in big data, artificial intelligence, information retrieval and knowledge discovery areas to place them in a strong position when applying for academic and industrial jobs.

这项工作的一个目标是将联合收割机搜索、知识图、问题回答和推荐的认知搜索异构语料库。提出的认知搜索架构是为了结合联合收割机强大的索引技术与先进的自然语言处理能力和机器学习算法，以建立一个越来越深的知识语料库，从其中提供相关信息给用户。本研究工作的三个主题是分别提出了数据驱动的知识图、问答系统和推荐系统。15名学生，包括研究生和本科生，将接受培训。一个博士每个项目将由一名本科生和一名硕士生完成，本科生主要将搜索引擎作为他们第四年的项目或论文。所有的方法都将在真实的数据集上进行评估，如Cuberweb 12-B13，Amazon，微博，Yelp，SemEval等。原型将在三个主题的每个阶段开发。 这项研究的结果将对加拿大，国际企业和政府机构有用。拟议的框架可供金融业使用（例如，道明银行和股票市场）、医疗保健（例如，海考会）、政府机构（例如，加拿大统计局）和技术公司（例如，Amazon和Google）。我们的计划将在大数据，人工智能，信息检索和知识发现领域培训学生，使他们在申请学术和工业工作时处于有利地位。