面向不一致用户评价准则的在线服务推荐

客观上,用户的评价准则是由主观意识决定的,用户之间的评价准则不同导致多个用户对同一服务的评分不具备可比较性,不考虑不同用户评分的不可比较性所获得的服务推荐将难以满足用户个性偏好及其真实需求。为此,提出一种面向不一致用户评价准则的在线服务推荐方法,考虑用户偏好不一致时用户对在线服务的偏好关系,以偏好关系计算用户之间的相似度,并以此获得在线服务推荐结果。首先以用户-服务评分矩阵为基础建立用户对服务的偏好关系,其次根据偏好关系计算用户之间的相似度,然后以用户相似度为基础对用户未评分的服务进行评分预测,最后以预测评分的排序结果作为推荐结果。与经典的协同过滤推荐方法的比较实验,验证了本方法的有效性。实验表明,本方法获得的推荐结果能满足大多数用户的服务偏好,同时获得了比经典的协同过滤推荐方法更好的准确率。     

用户群体满意度最大化的Top-k在线服务评价

考虑用户评价准则不一致的在线服务评价通常以服务的完整排序作为评价结果,而不是选择出使用户群体满意度最大的Top-k在线服务集合,使评价结果难以满足Top-k在线服务评价场景的合理性和公平性需求.为此,提出了一种用户群体满意度最大化的Top-k在线服务评价方法.该方法首先定义用户群体满意度指标,以衡量选择的k个在线服务的合理性;其次,考虑用户评价准则不一致及用户偏好信息不完整的情况,采用Borda规则将用户对在线服务的偏好关系构造为用户-服务满意度矩阵;然后借鉴Monroe比例代表思想,将Top-k在线服务评价问题建模为寻找最大化用户群体满意度的在线服务集合的优化问题;最后采用贪心算法对该优化问题进行求解,将得到的在线服务集合作为Top-k评价结果.通过理论分析和实验验证了该方法的合理性和有效性.理论分析表明,该方法满足Top-k在线服务评价所需的比例代表性和公平性.同时,实验结果也表明,该方法能够在合理的时间内获得接近用户群体满意度理想上界的评价结果,可以有效地辅助用户群体做出正确的服务选择决策.另外,该方法还可以在用户偏好不完整的情况下实现Top-k在线服务评价.     

一种QoS感知的Web服务群体评价方法

传统QoS感知的Web服务选择方法需要用户给出精确权重以表达其对QoS属性的偏好。为解决用户无明确偏好情况下的服务选择问题,提出了一种QoS感知的Web服务群体评价方法。对Web服务群体评价的需求进行了详细的分析,以有明确QoS偏好的用户对服务的历史选择结果为基础,利用群决策的Condorcet和Borda函数集结这些结果得到服务的群体评价,从而使得无明确QoS偏好用户可以参考该评价结果进行服务选择。理论分析和仿真实验验证了该服务群体评价方法的合理性和有效性。     

利用属性条件偏好推理的在线服务群体选择

在线服务群体选择能够在多人社交活动中提供适合群体的活动方案.在线服务群体选择中，如不考虑用户的服务属性偏好，则难以对未体验服务的偏好进行推理，导致对未体验服务的预测偏好与用户真实偏好存在差异，使得聚合个体用户偏好得到的群体选择结果难以真实反映群体对服务的偏好.为此，提出一种利用属性条件偏好推理的在线服务群体选择方法.首先根据用户对服务属性的偏好建立条件偏好表和属性之间的依赖关系；然后根据条件偏好网络(Condition Preference networks, CP-nets)的性质推理偏好导出图，偏好导出图的拓扑排序对应用户对服务属性值组合的偏好关系，得到个体用户的服务选择结果；最后使用社会选择函数Ranked Pairs对个体用户的服务选择结果进行聚合，得到群体的服务选择结果.通过实验验证了该方法的合理性和有效性.实验结果表明方法得到的群体结果与个体用户选择结果的平均nDCG(Normalized Discounted Cumulative Gain)比对比方法分别高11.4%、2.2%和11.1%,方法还满足多数准则、孔多赛性、抗操纵性等性质.     

基于社会选择理论的在线服务评价

用户评价标准不一致和偏好不一致导致网络空间中的在线服务之间不具备公正的可比较性,从而用户难以选择到满意的在线服务,因此,提出了基于社会选择理论计算在线服务优劣的排序方法。首先,根据用户给出的用户-服务评价矩阵构建群体偏好矩阵;然后,基于群体偏好矩阵和Kemeny社会选择函数构建0-1整数规划模型;最后,通过求解该模型可得到服务的最优排序结果。该方法聚合个体偏好为群体偏好,决策符合群体大多数人的偏好且与个体偏好保持最大的一致性。通过理论分析和实验验证了该方法的合理性和有效性。实验结果表明,该方法能有效地解决在线服务之间的不可比较性问题,实现在线服务的优劣排序,并可以有效抵制推荐攻击,具有较强的抗操纵性。     

类型相似性与品味影响力的推荐系统评分预测

电影推荐算法依靠计算用户的偏好差异进行推荐,以解决互联网时代的信息过载问题.传统协同过滤等推荐算法主要基于用户间对相同电影的评分差异计算用户偏好的相似性.这类方法忽视了用户的评分行为是一种实际上的选择行为,即便评分不高也体现出用户对该类型电影的兴趣.针对这一问题,本文设计了基于电影类型标签选择概率的用户间相似性计算方法,并建立了以用户为节点,以用户之间的相似性为边的推荐系统的复杂网络模型,并根据上述网络拓扑结构中的节点中心性数据,进一步设计了平衡用户品味影响力函数,调整了用户协同偏好的结果,提出了基于用户偏好相似性和用户品味影响力的电影评分预测方法.在MovieLens数据集上的实验结果表明,本文提出的算法与几种典型的现有方法相比较,可以有效的度量用户偏好的相似性以及抵消用户大众化品味影响力被高估在评分预测中带来的负面影响,与现有算法相比预测误差平均降低了2%至5%.     

基于深度信念网与隐变量模型的用户偏好建模

从高维、稀疏的用户评分数据中构建用户偏好模型,存在迭代计算复杂度高、中间结果规模大和难以实现有效推理等问题。为此,提出一种基于深度信念网(DBN)和贝叶斯网(BN)的用户偏好建模方法。采用DBN对评分数据进行分类,用隐变量表示不能直接观测到的用户偏好,利用含隐变量的BN描述评分数据中蕴含的相关属性间的依赖关系及其不确定性。在MovieLens和大众点评数据集上的实验结果表明,该方法能够有效描述评分数据中与用户偏好相关的各属性间的依赖关系,其精确率和执行效率均高于隐变量模型。     

利用Fallback的在线服务信誉防控制机制

在线服务信誉是若干服务信用行为累积的结果,对于在线服务选择具有重要的作用.信誉系统管理者为获取不当利益,可通过删除、增加用户或服务进行控制以达到操纵服务信誉的目的.为此提出利用Fallback的在线服务信誉防控制机制.首先获取所有用户对在线服务的序数偏好集合;其次根据所有用户的序数偏好得到满足Fallback绝对多数阈...     

支持QoS和用户需求的Web服务选择模糊算法*

基于QoS的Web服务选择能为用户提供更好的服务质量,而基于用户对QoS的需求和偏好的模糊性来进行Web服务选择是其中的一种方法。如何通过这种方法合理地获得最优候选服务集,是Web服务架构亟待解决的一个问题。本文提出了一种支持QoS和用户需求的Web服务选择模糊算法,该算法将用户对QoS的需求及偏好的语言描述进行去模糊化,然后选择出最符合用户需求的服务集。通过对模拟的服务质量数据进行实验,验证了该算法的优越性和有效性。     

基于用户偏好动态变化的协同过滤推荐

传统的协同过滤方法利用用户评分数据来生成推荐,没有考虑评价时间和项目类别等其他信息,影响了系统推荐的质量。本文提出一种基于用户偏好动态变化的个性化推荐模型,该方法在基于项目类别的基础上,为用户评分时间距离现在较近、较远和周期性评分分别赋予不同的权重。从MovieLens数据集的实验结果表明,该方法消弱了历史短期偏好对推荐质量的影响,准确地反映了用户偏好的动态变化,有效地提高了推荐的准确性。     

An intelligent web service group-based recommendation system for long-term composition

A modern model long-term composed service (LCS) with a group recommendation system has an indefinite lifespan. An LCS is used as a long-term business goal, and for a business committed to its customers, support will be provided to customers enabling them to book, e.g. an automotive service through online web services by providing information that the LCS then uses to offer more support. However, identifying the exact service to meet the user requirement is essential. Service composition has been identified as the key task in achieving various QoS performances. There exist various approaches that involve service composition according to the throughput and popularity. However, they fail to achieve the expected performance. Towards improving the performance of the LCS, a novel LCS that is based on the user queries of a group of persons is developed to give the best business services based on previous travel details and services. The method carries out service selection and composition according to the ratings provided by users towards any service. Additionally, the method considers the user-to-service rating and service-to-service rating, which are measured according to the coupling quality. Therefore, the proposed novel LCS provides better services based on the user ratings for particular business queries. The method ranks the services according to the rating values to perform service composition, with consideration of the detection of similar user groups and utilization of the rating values in service selection. We aim to propose a novel LCS work based on group ratings and a group of services. This work is intended to reduce the time complexity of changes in the LCS network using the group recommendation system.

     

Minimal functional routes in directed graphs with dependent edges

Graphs are mathematical structures used to model a set of objects and the relations between them. One of the basic concepts of graph theory, the path, has wide real‐world applications. In classic graph models, edges ending at a node are assumed to be independent. However, many real graphs/networks can only be correctly described by considering a dependency among nodes or edges. Paths in such graphs may not be functional if the conditional dependency is ignored. In this study, we investigate the routing problem in directed graphs with dependent edges represented by general graph models as alternatives to hypergraphs. We define a minimal functional route (MFR) as a minimal set of nodes and edges that can independently perform information transfer between two given nodes, and formulate the determination of MFRs as a graph search problem. A depth‐first‐search (DFS) top‐down algorithm, an iterative integer linear programming (ILP) bottom‐up algorithm, and a subgraph‐growing bottom‐up algorithm are devised subsequently to solve this problem. Numerical experiments verify the effectiveness of the algorithms. The defined MFR problem and the proposed algorithms are expected to find many practical applications.     

事件社交网中基于有向标签图及用户反馈的活动推荐方法

由于基于事件的社交网络（EBSN）中的活动具有时效性,传统社交网络推荐算法无法适用于EBSN。此外,大多数算法忽略了能影响后续推荐质量的前用户是否接受活动的反馈意见。为此,提出一种EBSN中基于有向标签图及用户反馈的活动推荐方法。首先,将EBSN抽象为有向标签图,并抽取图节点及边的属性特征信息,构建有向图结构特征（DGSF）索引,该索引由节点属性特征索引、有向边属性特征索引以及时间特征索引构成,利用该索引对节点及边进行初次过滤。其次,提出基于DGSF索引的多属性候选集过滤策略,利用时间、节点的出入度、标签类型等特征的限制,实现对查询图候选集的进一步剪枝,避免冗余计算。然后,提出一种具有用户反馈的改进UCB（Upper Confidence Bound）活动推荐算法——EN_UCB,通过引入弹性网回归,根据多影响因素计算用户对活动的兴趣值,为用户推荐兴趣值高的活动,同时接收用户是否接受该活动的反馈,以优化后续用户的推荐。大量实验结果表明,EN_UCB算法的接受率高于TS（Thompson Sampling）、UCB以及eGreedy算法,遗憾率远远低于TS和eGreedy算法,且运行效率高于TS、UCB以及eGreedy算法,活动数越大,优势越明显。所提算法能有效实现EBSN上的在线活动推荐。     

一种基于半环理论的可信性评估模型

在网络安全领域,可信指的是参加各种协议的各个实体之间关系的集合,这些关系是建立在实体在某个协议之上进行相互操作的行为之上的,为了加强网络的安全性,评估结点的可信性是非常重要的。讨论了对可信事件的评估：首先介绍了可信的相关概念和相关的特性;接着评估过程被建模成一个在有向图寻找最短路径的问题,在该有向图中结点表示实体或者用户,边表示可信关系,通过使用半环理论,建立了一个基于半环的可信性评估模型TD-SEMIRING,在两个以前没有进行相互操作的实体之间建立间接的可信关系,介绍了这个模型在路由选择当中的应用;最后,通过仿真实验,验证、分析了该模型的有效性。     

Stochastic graph as a model for social networks

Social networks are usually modeled and represented as deterministic graphs with a set of nodes as users and edges as connection between users of networks. Due to the uncertain and dynamic nature of user behavior and human activities in social networks, their structural and behavioral parameters are time varying parameters and for this reason using deterministic graphs for modeling and analysis of behavior of users may not be appropriate. In this paper, we propose that stochastic graphs, in which weights associated with edges are random variables, may be a better candidate as a graph model for social network analysis. Thus, we first propose generalization of some network measures for stochastic graphs and then propose six learning automata based algorithms for calculating these measures under the situation that the probability distribution functions of the edge weights of the graph are unknown. Simulations on different synthetic stochastic graphs for calculating the network measures using the proposed algorithms show that in order to obtain good estimates for the network measures, the required number of samples taken from edges of the graph is significantly lower than that of standard sampling method aims to analysis of human behavior in online social networks.     

The study of trust vector based trust rating aggregation in service-oriented environments

In most existing studies on trust evaluation, a single trust value is aggregated from the ratings given to previous services of a service provider, to indicate his/her current trust level. Such a mechanism is useful but may not be able to depict the trust features of a service provider well under certain circumstances. Alternatively, a complete set of trust ratings can be transferred to a service client for local trust evaluation. However, this incurs a big overhead in communication, since the rating dataset is usually in large scale covering a long service history. The third option is to generate a small set of data that should represent well the large set of trust ratings of a long time period. In the literature, a trust vector approach has been proposed, with which a trust vector of three values resulting from a computed regression line can represent a set of ratings distributed within a time interval (e.g., a week or a month, etc.). However, the computed trust vector can represent the set of ratings well only if these ratings imply consistent trust trend changes and are all very close to the obtained regression line. In a more general case with trust ratings for a long service history, multiple time intervals have to be determined, within each of which a trust vector can be obtained and can represent well all the corresponding ratings. Hence, a small set of data can represent well a large set of trust ratings with well preserved trust features. This is significant for large-scale trust rating transmission, trust evaluation and trust management. In this paper, we propose one greedy and two optimal multiple time interval (MTI) analysis algorithms. We also have studied the properties of our proposed algorithms analytically and empirically. These studies can illustrate that our algorithms can return a small set of MTI to represent a large set of trust ratings and preserve well the trust features.     

基于图卷积神经网络的在线社区行为预测

在线知识社区中,问题的回答可以看作多个回答者用户(领域专家)之间的协作行为。协作行为在知识社区中通常是大规模地发生,协作行为预测对在线社交中领域专家的推荐有重要意义。基于在线知识社区中回答者用户之间的协作行为,构建以领域专家为节点,以他们之间的协作回答关系为边的协作网络。由于协作行为网络的构建与社交关系网络的构建上结构的相似性,可以将协作行为预测构建为协作网络中的链接预测问题。通过构建基于图卷积神经网络的链接预测模型,对在线知识社区中回答者用户的协作行为进行预测。基于“知乎”数据集的实验验证,与其他经典的预测方法进行比较时,发现提出的方法能够更加有效地预测在线知识社区中回答者用户之间的协作行为。     

C语言的别名分析方法研究

别名分析对于数据流分析、程序优化和分析工具的实现非常重要.文章提出了一种需求驱动,流非敏感的分析算法来解决指针别名问题.通过构造程序表达式图(PEG)把指针别名问题转化成判断两个指针节点是否是联通的问题,它不同于传统的别名分析方法,它不需要构造别名集合和对其求交集,所以提高了分析指针别名的效率.