基于本体论的大数据下用户需求表征

1)深圳大学管理学院,广东深圳 518060; 2)深圳大学人因工程研究所,广东深圳 518060

自然语言处理; 大数据; 本体论; 用户需求表征; 新产品开发; 模糊需求

Ontology-based user requirements representation in the context of big data
Chen Xingyu1, Huang Junwen1, Zhou Zhan1, and Qu Xingda2

Chen Xingyu1, Huang Junwen1, Zhou Zhan1, and Qu Xingda21)College of Management, Shenzhen University, Shenzhen 518060, Guangdong Province, P.R.China2)Institute of Human Factors and Ergonomics, Shenzhen University, Shenzhen 518060, Guangdong Province, P.R.China

natural language processing; big data; ontology; user requirement representation; new product development; fuzzy requirements

DOI: 10.3724/SP.J.1249.2017.02173

备注

针对目前用户需求提取与表征研究中无法清晰表征大规模模糊需求的局限性,提出一种基于本体论的用户需求表征方法,通过自然语义处理、智能机器学习等人工智能算法,将初始数据生成、需求本体生成以及需求表征库生成整合成一套完整用户需求表征的科学指导方法.这一表征方法将本体论与大数据处理相结合,能在海量网络数据中,准确提取用户需求中的概念、分类及非分类关系,清晰地表征用户需求,特别是用户的模糊需求.这些需求可具象为产品特征,用以建立相应的表征库,为新产品开发与创新提供有效支持.

Representation of user requirements is essential for product innovation. However, traditional market survey methods seem to fail to represent user requirements effectively in current big data context. We propose a novel ontology-based method to overcome limitations of existing research on the extraction and representation of users' requirements.This method integrates initial data generation, requirement ontology generation and demand characterization into a complete set of scientific guidance for user demand characterization using natural semantic processing, intelligent machine learning and other artificial intelligence algorithms. It uses ontology and big data processing techniques to extract concepts, taxonomic and non-taxonomic relations from huge raw internet data, and further helps identify user's requirements, especially fuzzy requirements. The identified user requirements can be embodied as product features, which helps build product feature representation library to provide support for new product development and innovation.

·