云模型和多特征的高校读者借阅偏好不确定性图书推荐研究

Research of Books Recommendation of Borrow Preference Uncertainty in University Readers Based on Cloud Model and Multi-feature

利用云模型表示自然界中模糊性、随机性等不确定性优势,提出云模型和读者多特征的借阅偏好不确定性。计算读者专业、性别、年级加权相似度,利用逆向云算法计算以云的期望、熵、超熵来表示的读者借还时间间隔偏好,再计算读者基于云的相似度。结合读者多特征相似度、云相似度,向读者推荐存在复本的图书,并通过实验验证算法的有效性。

The paper, taking advantages of cloud model in expressing uncertainty of fuzziness and randomness in nature, proposes that there are uncertainties in cloud model and readers＇ multi - featured books preference. The paper computes the weighted similarity of readers＇ majors, genders and grades, uses the backward cloud algorithm to compute readers＇ preference of borrowreturn time interval that is expressed in cloud expectation, entropy and hyper entropy, and then computes the readers＇ cloud similarity. Combining multi - featured similarity of readers with cloud similarity, the paper finally recommends books to readers with copied ones and proves the validity of the algorithm by experiments.