SBIR Phase II: Hybrid Question Answering Combining a Search Index with an RDF Store

SBIR 第二阶段：将搜索索引与 RDF 存储相结合的混合问答

• 批准号: 1230248
• 负责人: Marta Tatu
• 金额: $ 46.7万
• 依托单位: Lymba Corporation
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2012
• 资助国家: 美国
• 起止时间: 2012-09-01 至 2016-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1230248&HistoricalAwards=false
• 关键词: SBIR; Phase; II; Hybrid; Question

This Small Business Innovation Research (SBIR) Phase II project will provide integrated, seamless access to structured and unstructured information. At present there is no easy way to perform a single federated search across structured databases and unstructured text documents, and complex applications over diverse data sources require considerable time and effort to develop. The multilingual Hybrid Question Answering (HQA) engine will change the way people access heterogeneous data by answering complex questions over semantic models from the database and free text from the search indexes of product literature, call center data, social media, etc. HQA will answer a broad range of questions including factoid, procedural, explanation and scenario-matching. Users will also be able to discover facts, relationships among events, and sentiments; to identify trends and enable predictions. Finally, to ensure high precision, HQA will process information in its native form for 11 languages, including English. A minimum score of 70% MRR is expected for the English HQA system and 60% MRR for the other 10 languages.The broader impact/commercial potential of this project spans several key areas. The HQA engine will surpass existing Questioning Answering technology by answering a wide range of complex questions with information found in both structured and unstructured data sources. Social media-based applications such as trend finding, sentiment detection, and predictive analysis operating in international markets will greatly benefit from the availability of the multi-lingual HQA system.Customer Relation Management (CRM), Business Intelligence and Social Media are among the primary commercial applications benefiting from the development of the HQA engine. The impact of this technology will result in businesses being able to develop more personalized relationships with customers as a result of a better understanding of customer needs and motivations. Traditional business relationships will become more responsive, resulting in greater understanding and trust, which will mitigate destructive, wide swings in financial markets caused by poor predictive models.

这个小型企业创新研究（SBIR）第二阶段项目将提供对结构化和非结构化信息的集成、无缝访问。目前，还没有简单的方法来执行跨结构化数据库和非结构化文本文档的单一联合搜索，并且在不同数据源上的复杂应用程序需要大量的时间和精力来开发。多语言混合问题查询（HQA）引擎将通过回答来自数据库的语义模型和来自产品文献，呼叫中心数据，社交媒体等搜索索引的自由文本的复杂问题来改变人们访问异构数据的方式。HQA将回答广泛的问题，包括事实，程序，解释和匹配。用户还将能够发现事实、事件之间的关系和情绪;识别趋势并进行预测。最后，为了确保高精度，HQA将以包括英语在内的11种语言的本地形式处理信息。预计英语HQA系统的MRR最低得分为70%，其他10种语言的MRR最低得分为60%。该项目的更广泛影响/商业潜力跨越几个关键领域。HQA引擎将通过使用结构化和非结构化数据源中的信息回答各种复杂问题，超越现有的Questioning Questioning技术。基于社交媒体的应用程序，如趋势发现，情绪检测和预测分析在国际市场上运行将大大受益于多语言HQA系统的可用性。客户关系管理（CRM），商业智能和社交媒体是受益于HQA引擎开发的主要商业应用程序之一。这项技术的影响将使企业能够更好地了解客户的需求和动机，从而与客户建立更加个性化的关系。传统的商业关系将变得更加敏感，从而产生更大的理解和信任，这将减轻由于预测模型不佳而导致的金融市场的破坏性和大幅波动。