【摘要】 图书采访是一种建立在不确定的需求预测基础上的决策活动,无论传统的主观定性决策采访,还是利用决策工具进行量化决策采访,面对读者和图书这两类具有较大差异决策对象和大量数据处理,都不能取得令人满意的效果。采用人工智能技术,模仿人类思维方式,在分析每个读者内在的知识需求特征基础上,根据读者与图书馆互动数据,对读者需求做出更加准确的预测;同时,用馆藏数据、历次采访数据、采访书单数据对人工智能不断训练和学习,逐渐形成与人类认知接近的馆藏特色、采访偏好、图书评价的数据。最后通过科学决策工具,取得最优化的图书采访效果。更多还原

【Abstract】 Book acquisition is a kind of decision-making activity based on uncertain demand forecasting. Regardless of the traditional subjective qualitative decision interview,or using decision making tool for quantitative decision interview,in the face of readers and books, the two types of different decision-making objects, and a large number of data processing,satisfactory effect cannot be obtained. Based on the analysis of characteristics of knowledge within each reader demand and the interaction data between the reader and library,this study uses AI technology to imitate human mode of thinking and make more accurate predictions for the reader needs,meanwhile uses collection data,previous acquisition data and acquisition booklist data for AI training and learning constantly, forms the collection features, interview preference and book evaluation data which are closer to human cognition. Finally through scientific decision-making tools,the result of the book acquisition is optimized.更多还原