Semantic information extraction for creating an expertise knowledge base

用于创建专业知识库的语义信息提取

• 批准号: 478332-2015
• 负责人: Lyons, Kelly
• 金额: $ 1.82万
• 依托单位: University of Toronto
• 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=579420
• 关键词: Semantic; information; extraction; creating; expertise

Knowledge workers comprise a large proportion of the workforce today in organizations. More often than not, they work in the context of large-scale expert networks, spanning different disciplines and activities. This type of networked interdisciplinary work is motivated by the belief that innovation is enhanced through broad collaboration among experts and partners who join forces to create value through new products, services and ideas. It is critical for organizations to build an innovation pipeline and culture in order to continually develop new ideas and identify opportunities for the development of intellectual property. An important opportunity for improving the innovation process is to help employees recognize novel ideas and communicate them to the experts who can help develop and disclose the ideas. However, finding people with the right expertise to help achieve a given goal is not trivial even when an expertise directory exists in the organization. In this project, we will combine information extraction (IE) and novel human-computation techniques to develop a platform for the creation, curation, and exploration of a semantically linked expertise knowledge base. Our hybrid system will mine unstructured and semi-structured data to extract information about ideas and people involved in innovation activities as well as their explicit and implicit roles and expertise. Discovered entities, and entity types will be linked together within the extracted dataset (intra-linkage) and externally with existing linked datasets such as Freebase and openCyc (inter-linkage). We will develop innovative methods and tools for curating the initial repository and extracting semantic entities and relations among them. The curation toolkit will use unique and original incentive-based interactive and collaborative human-computation techniques. Our platform will enable people to find expertise and individuals or groups with similar or complementary ideas who can be brought together to collaborate on the innovation process.

知识工作者构成了当今组织中劳动力的很大一部分。他们往往在跨越不同学科和活动的大规模专家网络中开展工作。这种跨学科的网络化工作的动机是相信通过专家和合作伙伴之间的广泛合作，通过新产品，服务和想法创造价值，从而促进创新。组织必须建立创新渠道和文化，以便不断开发新的想法，并确定知识产权发展的机会。改进创新过程的一个重要机会是帮助员工识别新的想法，并将其传达给可以帮助开发和披露这些想法的专家。然而，找到具有正确专业知识的人来帮助 即使组织中存在专门知识目录，实现给定目标也不是微不足道的。 在这个项目中，我们将结合联合收割机信息提取（IE）和新颖的人类计算技术，开发一个平台，用于创建，策展和探索一个语义链接的专业知识库。我们的混合系统将挖掘非结构化和半结构化数据，以提取有关创新活动中的想法和人员的信息，以及他们明确和隐含的角色和专业知识。发现的实体和实体类型将在提取的数据集内链接在一起（内部链接），并在外部与现有的链接数据集（如Freebase和openCyc）链接在一起（内部链接）。我们将开发创新的方法和工具来管理初始存储库，并提取语义实体及其之间的关系。策展工具包将使用独特和原始的基于激励的交互和协作人类计算技术。我们的平台将使人们能够找到专业知识和个人或团体与类似或 互补的想法，谁可以聚集在一起，在创新过程中进行合作。