基于协同过滤算法的高校图书馆图书推荐系统研究

基于协同过滤算法是个性化推荐系统中最基本、最常用的一种,而传统的基于用户的协同过滤算法存在着用户冷启动和高推荐率等问题。笔者针对目前图书推荐协同过滤算法运用中存在的问题,并根据高校图书馆的实际情况,提出相应的改进措施。笔者利用学生网络日志和读者借阅记录对用户进行相似性分析,有效解决了用户“冷启动”的问题,提高了推荐准确率。笔者利用时间衰减法建立用户兴趣模型,以便更好地关注用户的短期借阅行为,提高推荐准确率。与传统用户协同过滤方法相比,笔者提出了一种改进的用户协同过滤方法。结果表明,改进后的用户协同过滤方法具有更好的性能。     

电子商务中Web页面个性化推荐系统的架构

电子商务中实现个性化推荐意味着一个用户访问Web站点时能够得到个性化的服务,网站根据用户的聚类特征,向用户在线推荐一些用户可能比较感兴趣的页面。本文给出了电子商务个性化推荐的系统结构,分析了在Web日志挖掘中应用协同过滤技术,讨论了Web页面的个性化推荐过程和推荐算法。推荐算法综合考虑了用户聚类中页面的权值和用户对页面的平均评价值两个推荐因素,实现在线页面的推荐。     

邮件营销个性化推荐方法

在电子商务中,邮件营销成为企业竞争的另一种形式。但是因缺乏明确的发送目标,许多带有营销目的的电子邮件被视同为垃圾邮件,起不到营销效果。笔者在分析当前推荐技术中各种算法的优缺点的基础上,提出了一种基于用户行为偏好的动态序列挖掘方法。这种算法根据用户购买信息,运用数据挖掘技术,将隐含其中的用户特征、偏好等重要信息挖掘出来,从而预测用户最感兴趣的产品,有针对性地为用户推荐个性化的信息。实验证明,这种方法有效。     

社交电商中融合信任和声誉的图神经网络推荐研究

社交电商可依据用户间的社交关系为用户提供感兴趣的商品或服务。现有研究多基于社会信任或社会声誉进行推荐，却忽略了信任与声誉间的相互作用，导致推荐效果欠理想。针对以上问题，本文提出了一种融合信任(Trust)和社会声誉(Social Reputation)的图神经网络推荐算法(TSR-GM)，采用社会声誉来深度刻画用户关系在推荐系统中的作用，利用社交网络中用户被信任程度对用户声誉进行排名，以图神经网络量化整合用户信任与声誉，并将结合后的新矩阵不断校正以获取更准确的用户信任，以此对矩阵分解后得到的新评分模型更新，最终得到更准确度量的预测评分矩阵。运用Epinions数据集开展的相关实验表明：与同类方法比，TSR-GM算法对提高推荐精度有较好效果。     

电子商务协同过滤稀疏性研究:一个分类视角

协同过滤是目前电子商务推荐系统中广泛使用的、最成功的推荐算法,但面临用户评分数据稀疏性问题的挑战.在介绍用户偏好数据获取途径的基础上,将稀疏性改善技术归纳为六类,包括设定缺省值、结合基于内容的过滤、降维、图论方法、基于项目评分预测以及增加用户.系统交互,重点评述了各类算法的研究情况并时六类技术进行了比较,最后探讨了该领域的未来研究方向.     

基于用户画像的个性化图书推荐研究

目的/意义:运用个性化推荐服务,满足高校读者的个性化需求。方法/过程:首先分析读者身份信息和历史借阅行为信息数据,然后创建图书馆读者的用户画像标签模型,最后结合个性化推荐算法构建智慧阅读推荐系统。结论:图书馆可以通过用户画像标签快速了解读者群体的兴趣方向。     

基于遗忘函数和领域最近邻的混合推荐研究

基于内容过滤和协同过滤是两大最为经典的推荐算法,但基于内容过滤存在新用户问题,没有考虑用户兴趣变化对推荐质量的影响,协同过滤则面临严峻的数据稀疏性和冷启动的挑战.针对这些,提出混合推荐算法:基于非线性逐步遗忘函数建立用户兴趣模型,预测用户未评价商品评分;引入"领域最近邻"处理方法查找目标用户的最近邻,预测未评价商品评分,以此为基础做出推荐.实验结果表明,本文方法能有效提高推荐质量.     

团购商品的个性化协同过滤推荐方法

本文提出结合云模型和改进均方差的协同过滤推荐方法,为团购网站的用户提供个性化推荐。该方法首先通过已有的用户日志数据,模拟用户对商品的评分,构建用户-商品评分矩阵。在此基础上,通过云模型和改进MSD方法模拟商品间的相似度,运用基于物品的协同过滤方法,预测出用户对商品的评分,并为目标用户生成个性化推荐列表。最后,通过实验证明该推荐方法的有效性。     

电子商务个性化文档推荐技术研究

电子商务个性化文档推荐同时具有电子商务推荐系统以及文档信息处理的特点,由此本文首先概括了目前电子商务推荐以及文档信息处理的主要技术,多种推荐技术的有效集成会提高推荐的精度与个性化,也是未来电子商务中个性化文档推荐的发展趋势。其次阐述了几种重要的用户兴趣获取技术及其面临的主要问题。最后分析了电子商务个性化文档推荐技术面临的主要问题及未来发展方向,旨在挖掘用户潜在兴趣,提高文档推荐个性化和提高推荐质量。     

基于上下文感知的组推荐系统框架设计

在信息服务领域,个性化推荐系统为用户解决了"信息过载"的问题。本文首先分析了以往推荐系统的优缺点,在此基础上引入上下文和组推荐,将两者结合起来,形成一个新的组推荐框架,然后详细阐述了此框架的各个模块。     

Developing expert systems for management applications

The Alvey Report has resulted in a growing interest in the UK in ‘expert systems’. It is fairly generally accepted, at least in the UK, that such systems function in a particular type of way, i.e. they arrive at decisions through a process of rule based inference. It is suggested that it may be more fruitful to regard rule based inference as one approach to the construction of expert systems, and that proven techniques of operational research may well be more useful in constructing other types of expert system. Alternative applications of expert systems are derived on the basis of a broader definition of an expert system in terms of what it does rather than how it does it. A parallel is drawn between these applications and some typical concerns of business research. It is suggested that a useful aid in identifying promising business applications of expert systems is to set up four ‘dimensions’ along which different types of system differ. Examples are given of where other techniques might conceivably be useful in applications of expert systems.     

决策单元交叉效率的自适应群评价方法

传统交叉效率评价方法因决策单元偏好权重不唯一而难以操作,因交叉效率有效性分值平均化集结而难以被接受。目前的学者通常围绕决策单元指标权重的确定性分配方法、交叉效率有效性分值的去平均化集结等分别开展研究。本文将交叉效率评价方法中自评互评相结合的评价模式看作群决策过程,即每个决策单元既是一个被评对象,又是一个决策"专家",提出了一种决策单元交叉效率的自适应群评价方法,将决策单元偏好权重的确定和交叉效率有效性分值的去平均化集结作为同一个决策过程,根据每个决策单元的评价结果与群体评价结果的接近程度,同步迭代调整决策单元的"专家"权重和决策单元自评产生的、并提供给其他被评价决策单元的一组确定的偏好指标权重。实验验证与实例运用分析表明,该方法收敛效果良好,能得到客观稳定的决策单元交叉效率有效性分值及排序。     

基于个人知识地图的专家推荐

利用维基百科作为背景知识构建出专家个人的知识地图,从而直观量化地度量专家的知识构成和研究兴趣。在此基础上,提出了基于知识节点密度和最大公共子图的2种推荐算法,并且将这2种推荐算法和经典的推荐算法结合。最后,用一个真实的数据集合验证:这种基于个人知识地图的专家推荐是有效的。     

An Application of the AXIS Solution Framework to Multiple Objective Aggregate Production Planning

This paper provides data on the first application of a prototype of the AXIS solution framework. AXIS (algorithms combined with knowledge systems in an interactive sequence) is a framework for interactively combining structured algorithms that seek a best solution with knowledge-based expert systems that seek expert heuristic solutions. This paper tests the framework using an interactive multiple objective integer programming algorithm combined with heuristics taken from the domain of aggregate production planning. The results indicate the AXIS framework can be successful in generating high quality solutions, in vastly reduced solution times compared to the structured algorithms, at much lower costs compared to the expert heuristics working alone.     

推荐系统对消费者网购支出的影响研究

推荐系统能在电子商务中利用信息过滤技术为消费者推荐感兴趣的商品和服务。本文通过收集大量消费者网购调查问卷，问卷的信度和效度均符合数据分析的要求。首先分析了消费者的产品偏好、忠诚度等网购行为与其年龄、性别、地域等个人属性的相互关系，之后运用倾向得分匹配法（Propensity Score Matching，PSM）研究推荐系统对消费者网购支出的影响，同时使用工具变量法（（Instrumental Variable，IV）对PSM研究结果进行稳健性检验。结果显示使用推荐系统的消费者网购支出比未使用的消费者高出14.7%，网购支出与受教育程度和收入水平呈正相关、与年龄呈负相关，城市消费者和女性更愿意使用推荐系统；同时统计分析表明推荐效果受电子商务平台中社交关系、互补产品、店铺声誉等因素影响。研究结果对评估推荐系统的经济效益、增强消费者忠诚度和提高商家营销的精准性等方面起到了重要作用。     

An Integrated Approach for Modeling Sustainability Risks in Freight Transportation Systems

Integrating sustainability into freight transportation systems (FTSs) is a complex and challenging task due to the sheer number of inherent sustainability risks. Sustainability risks disrupt the economic, social and environmental objectives of freight operations and act as impediments in the development of sustainable freight transportation systems. The area of sustainability risk management is still unexplored and immature in the operational research domain. This study addresses these research gaps and contributes in a threefold manner. First, a total of 36 potential sustainability risks related to FTSs are identified and uniquely classified into seven categories using a rigourous approach. Second, the research proposes two prominent perspectives, namely, ontological and epistemological perspectives to understand risks and develops a novel framework for managing sustainability risks in FTSs. Third, a novel approach by integrating fuzzy evidential reasoning algorithm (FERA) with expected utility theory is developed to quantitatively model and profile sustainability risk for different risk preferences, namely, risk-averse, risk-neutral, and risk-taking scenarios. The proposed FERA is a flexible and robust approach, which transforms the experts’ inputs into belief structures and aggregates them using the evidence combination rule proposed in Dempster–Shafer theory to overcome the problem of imprecise results caused by average scoring in existing models. A sensitivity analysis is conducted to demonstrate the robustness of the proposed model. Unlike conventional perception, our study suggests that most of the high priority sustainability risks are behaviorally and socially induced rather than financially driven. The results provide significant managerial implications including a focus on skills development, and on social and behavioral dimensions while managing risks in FTSs.     

Risk assessment of outsourcing e-procurement services: integrating SWOT analysis with a modified ANP-based fuzzy inference system

With electronic procurement solutions becoming increasingly sophisticated, many firms opt to source these services from third-party providers, effectively transferring (outsourcing) significant responsibility to these services companies. This action, however, entails certain risks, which are oftentimes difficult to assess. To guide managerial practice and to advance academic inquiry in this domain, we identify e-procurement risk factors through a strengths, weaknesses, opportunities, threats analysis, grounded in transaction cost economics (TCE), and propose a risk assessment framework based on the opinions of a group of experts. The approach taken is that of a modified analytic network process methodology, combined with a fuzzy inference system, which is versatile enough to accept the expert opinions in different input formats (such as linguistic variables and ranges). The proposed framework has the capability to aggregate expert judgements’ to estimate risk likelihoods, risk severities and risk factor indices, and derive overall risk magnitudes. The multi-method approach was motivated and is illustrated by a real-life case study of an Indian manufacturing company currently in the process of contract renewal with its existing e-procurement service provider.     

Subjective machines: Probabilistic risk assessment based on deep learning of soft information

For several years machine learning methods have been proposed for risk classification. While machine learning methods have also been used for failure diagnosis and condition monitoring, to the best of our knowledge, these methods have not been used for probabilistic risk assessment. Probabilistic risk assessment is a subjective process. The problem of how well machine learning methods can emulate expert judgments is challenging. Expert judgments are based on mental shortcuts, heuristics, which are susceptible to biases. This paper presents a process for developing natural language-based probabilistic risk assessment models, applying deep learning algorithms to emulate experts’ quantified risk estimates. This allows the risk analyst to obtain an a priori risk assessment when there is limited information in the form of text and numeric data. Universal sentence embedding (USE) with gradient boosting regression (GBR) trees trained over limited structured data presented the most promising results. When we apply these models’ outputs to generate survival distributions for autonomous systems’ likelihood of loss with distance, we observe that for open water and ice shelf operating environments, the differences between the survival distributions generated by the machine learning algorithm and those generated by the experts are not statistically significant.     

Expert knowledge,expert systems and commercial interests

Experience to date in building expert systems has brought a general realization of the complexity of the effort required for producing systems capable of operating on ‘real-world’ problems. The most successful knowledge-based/expert systems built to date, e.g. DENDRAL, MYCIN, PROSPECTOR and R1 have demonstrated that a considerable investment in time and dedication on the part of systems designers and experts is required to create a fully operational system. The great majority of development projects attempting to harness this new programming technology do not extend beyond creation of a prototype system largely due to lack of necessary resources, i.e. time, money and know-how. The purpose of this paper is to discuss some of the more important issues associated with development of expert systems and to provide an overview of the commercial and industrial efforts of putting expert systems to work. This paper should be of interest to those who have gained their knowledge of expert systems from the reading of non-specialist publications and thus may have been exposed to somewhat over enthusiastic accounts of the subject.