Predicting program execution times by analyzing static and dynamic program paths

This paper describes a method to predict guaranteed and tight deterministic execution time bounds of a sequential program. The basic prediction technique is a static analysis based on simple timing schema for source-level language constructs, which gives accurate predictions in many cases. Using powerful user-provided information, dynamic path analysis refines looser predictions by eliminating infeasible paths and decomposing the possible execution behaviors in a pathwise manner. Overall prediction cost is scalable with respect to desired precision, controlling the amount of information provided. We introduce a formal path model for dynamic path analysis, where user execution information is represented by a set of program paths. With a well-defined practical high-level interface language, user information can be used in an easy and efficient way. We also introduce a method to verify given user information with known program verification techniques. Initial experiments with a timing tool show that safe and tight predictions are possible for a wide range of programs. The tool can also provide predictions for interesting subsets of program executions.This research was supported in part by the Office of Naval Research under grant number N00014-89-J-1040.     

基于剪枝技术的自适应PPM预测模型

建立有效的用户浏览预测模型,对用户的浏览行为进行准确的预测,是Web预取的关键。标准PPM预测模型由于存在存储复杂度高、执行效率低等缺点,影响了其推广和应用。文章基于剪枝技术,依据Zipf法则及Web对象访问特征对标准PPM预测模型进行预先剪枝和后剪枝,构造出一种自适应PPM预测模型。实验表明,该模型不仅能动态预测用户的Web浏览特征,而且在预测准确率和存储复杂度方面都有一定程度的提高。     

The impact of data obfuscation on the accuracy of collaborative filtering

Collaborative filtering (CF) is a widely-used technique for generating personalized recommendations. CF systems are typically based on a central storage of user profiles, i.e., the ratings given by users to items. Such centralized storage introduces potential privacy breach, since all the user profiles may be accessible by untrusted parties when breaking the access control of the centralized system. Hence, recent studies have focused on enhancing the privacy of CF users by distributing their user profiles across multiple repositories and obfuscating the user profiles to partially hide the actual user ratings. This work combines these two techniques and investigates the unavoidable side effect of data obfuscation: the reduction of the accuracy of the generated CF predictions. The evaluation, which was conducted using three different datasets, shows that considerable parts of the user profiles can be modified without observing a substantial decrease of the CF prediction accuracy. The evaluation also indicates what parts of the user profiles are required for generating accurate CF predictions. In addition, we conducted an exploratory user study that reveals positive attitude of users towards the data obfuscation.     

Self-tuning filters and predictors for two-dimensional systems Part 3: Prediction applications

The paper describes the practical uses of two-dimensional (2-D) fc-step-ahead self-tuning prediction algorithms. Two distinct application areas are considered. The first concerns direct prediction/forecasting, applied to data with a strong periodic (or 'seasonal’) component. The second concerns the prediction of data from spatial scanning sensors or spatial sensor arrays. In both cases, the original data is usually one-dimensional in nature. The contribution of the paper is to show how, by treating the data as if it were two-dimensional in nature, a vast improvement in the quality of predictions is obtained. Moreover, because the 2-D predictors are self-tuning, the algorithms require virtually no user intervention and no prior filtering or pre-processing of the data.     

Realistic data transfer scheduling with uncertainty

Next Generation Networks (NGNs) will be comprised of different access technologies. We are already seeing the emergence of mobile devices with the capability of connecting to heterogeneous networks with different capabilities and constraints. In addition, many bandwidth intensive applications have rather relaxed real-time constraints allowing for alternative scheduling mechanisms which can take into account user preferences, network characteristics as well as future network resource availability to better exploit network heterogeneity. The current approaches either simply react to changes, or assume that availability predictions are perfect.In this paper, we propose a scheduling scheme based on stochastic modeling to account for prediction errors. The scheme optimizes overall user utility gain considering imperfect predictions taken over realistic time intervals while catering for different applications’ needs. We use 180 days of real user data of many users to demonstrate that it consistently outperforms other non-stochastic and greedy approaches in typical networking environments.     

Prediction of Web Page Accesses by Proxy Server Log

As the population of web users grows, the variety of user behaviors on accessing information also grows, which has a great impact on the network utilization. Recently, many efforts have been made to analyze user behaviors on the WWW. In this paper, we represent user behaviors by sequences of consecutive web page accesses, derived from the access log of a proxy server. Moreover, the frequent sequences are discovered and organized as an index. Based on the index, we propose a scheme for predicting user requests and a proxy-based framework for prefetching web pages. We perform experiments on real data. The results show that our approach makes the predictions with a high degree of accuracy with little overhead. In the experiments, the best hit ratio of the prediction achieves 75.69%, while the longest time to make a prediction only requires 2.3 ms.     

Learning to Predict by the Methods of Temporal Differences

This article introduces a class of incremental learning procedures specialized for prediction – that is, for using past experience with an incompletely known system to predict its future behavior. Whereas conventional prediction-learning methods assign credit by means of the difference between predicted and actual outcomes, the new methods assign credit by means of the difference between temporally successive predictions. Although such temporal-difference methods have been used in Samuel's checker player, Holland's bucket brigade, and the author's Adaptive Heuristic Critic, they have remained poorly understood. Here we prove their convergence and optimality for special cases and relate them to supervised-learning methods. For most real-world prediction problems, temporal-difference methods require less memory and less peak computation than conventional methods and they produce more accurate predictions. We argue that most problems to which supervised learning is currently applied are really prediction problems of the sort to which temporal-difference methods can be applied to advantage.     

Effective methods for increasing aggregate diversity in recommender systems

In order to make a recommendation, a recommender system typically first predicts a user’s ratings for items and then recommends a list of items to the user which have high predicted ratings. Quality of predictions is measured by accuracy, that is, how close the predicted ratings are to actual ratings. On the other hand, quality of recommendation lists is evaluated from more than one perspective. Since accuracy of predicted ratings is not enough for customer satisfaction, metrics such as novelty, serendipity, and diversity are also used to measure the quality of the recommendation lists. Aggregate diversity is one of these metrics which measures the diversity of items across the recommendation lists of all users. Increasing aggregate diversity is important because it leads a more even distribution of items in the recommendation lists which prevents the long-tail problem. In this study, we propose two novel methods to increase aggregate diversity of a recommender system. The first method is a reranking approach which takes a ranked list of recommendations of a user and reranks it to increase aggregate diversity. While the reranking approach is applied after model generation as a wrapper the second method is applied in model generation phase which has the advantage of being more efficient in the generation of recommendation lists. We compare our methods with the well-known methods in the field and show the superiority of our methods using real-world datasets.     

协同过滤系统隐私保护和推荐准确性研究

针对协同过滤推荐系统在预测过程中容易泄漏用户概貌数据的问题,在不影响推荐准确性的前提下,提出一种用户数据混淆策略,使响应用户的评分数据在计算用户相似度之前被假数据代替,用户尽量少泄露(或不泄露)个人评分信息,进而实现用户隐私的保护。通过实验分析数据混淆策略对协同过滤推荐准确性的影响,证明该策略的有效性。     

Trust-aware recommendation based on heterogeneous multi-relational graphs fusion

Users’ trust relations have a significant influence on their choice towards different products. However, few recommendation or prediction algorithms both consider users’ social trust relations and item-related knowledge, which makes them difficult to cope with cold start and the data sparsity problems. In this paper, we propose a novel trust-ware recommendation method based on heterogeneous multi-relational graphs fusion, termed as T-MRGF. In contrast with other traditional methods, it fuses the user-related and item-related graphs with the user–item interaction graph and fully utilizes the high-level connections existing in heterogeneous graphs. Specifically, we first establish the user–user trust relation graph, user–item interaction graph and item–item knowledge graph, and the user feature and item feature, which have been obtained from the user–item graph, are used as the input of the user-related graph and the item-related graph respectively. The fusion is achieved through the cascade of feature vectors before and after feature propagation. In this way, the heterogeneous multi-relational graphs are fused for the feature propagation, which largely refines the user and item representation for model prediction. Simulation results show that the proposed method significantly improve the recommendation performance compared to the state-of-the-art KG-based algorithms both in accuracy and training efficiency.     

基于辅助特征的Web服务QoS预测

随着互联网以及Web服务技术的快速发展,相同功能的Web服务数量越来越多.在构建面向服务的应用时,服务质量（QoS）作为Web服务的非功能特性开始被越来越多的用户所重视.为了向用户推荐高质量的服务,首先我们需要对服务质量进行预测.现今有很多关于Web服务QoS预测的工作,这些研究大都关注在建模方法的优化上,忽视了辅助特征对于QoS预测的影响.着重分析辅助特征对于QoS预测的影响,例如服务类别和用户地理位置.为了实现此目标,基于因子分解机（Factorization Machines）设计并构建了一个统一的QoS预测架构,该架构可以灵活、方便地考虑进多个辅助特征.结合服务类别和用户地理位置这两类辅助特征,提出了一种QoS预测方法,并通过在真实数据上的实验证明了我们的方法的优越性.     

Evaluating Variable-Length Markov Chain Models for Analysis of User Web Navigation Sessions

Markov models have been widely used to represent and analyze user Web navigation data. In previous work, we have proposed a method to dynamically extend the order of a Markov chain model and a complimentary method for assessing the predictive power of such a variable-length Markov chain. Herein, we review these two methods and propose a novel method for measuring the ability of a variable-length Markov model to summarize user Web navigation sessions up to a given length. Although the summarization ability of a model is important to enable the identification of user navigation patterns, the ability to make predictions is important in order to foresee the next link choice of a user after following a given trail so as, for example, to personalize a Web site. We present an extensive experimental evaluation providing strong evidence that prediction accuracy increases linearly with summarization ability     

Backfilling Using System-Generated Predictions Rather than User Runtime Estimates

The most commonly used scheduling algorithm for parallel supercomputers is FCFS with backfilling, as originally introduced in the EASY scheduler. Backfilling means that short jobs are allowed to run ahead of their time provided they do not delay previously queued jobs (or at least the first queued job). However, predictions have not been incorporated into production schedulers, partially due to a misconception (that we resolve) claiming inaccuracy actually improves performance, but mainly because underprediction is technically unacceptable: users will not tolerate jobs being killed just because system predictions were too short. We solve this problem by divorcing kill-time from the runtime prediction and correcting predictions adaptively as needed if they are proved wrong. The end result is a surprisingly simple scheduler, which requires minimal deviations from current practices (e.g., using FCFS as the basis) and behaves exactly like EASY as far as users are concerned; nevertheless, it achieves significant improvements in performance, predictability, and accuracy. Notably, this is based on a very simple runtime predictor that just averages the runtimes of the last two jobs by the same user; counter intuitively, our results indicate that using recent data is more important than mining the history for similar jobs. All the techniques suggested in this paper can be used to enhance any backfilling algorithm and are not limited to EASY     

Sequence-based trust in collaborative filtering for document recommendation

Collaborative filtering (CF) recommender systems have emerged in various applications to support item recommendation, which solve the information-overload problem by suggesting items of interest to users. Recently, trust-based recommender systems have incorporated the trustworthiness of users into CF techniques to improve the quality of recommendation. They propose trust computation models to derive the trust values based on users' past ratings on items. A user is more trustworthy if s/he has contributed more accurate predictions than other users. Nevertheless, conventional trust-based CF methods do not address the issue of deriving the trust values based on users' various information needs on items over time. In knowledge-intensive environments, users usually have various information needs in accessing required documents over time, which forms a sequence of documents ordered according to their access time. We propose a sequence-based trust model to derive the trust values based on users' sequences of ratings on documents. The model considers two factors – time factor and document similarity – in computing the trustworthiness of users. The proposed model enhanced with the similarity of user profiles is incorporated into a standard collaborative filtering method to discover trustworthy neighbors for making predictions. The experiment result shows that the proposed model can improve the prediction accuracy of CF method in comparison with other trust-based recommender systems.     

From similarity perspective: a robust collaborative filtering approach for service recommendations

Collaborative filtering (CF) is a technique commonly used for personalized recommendation and Web service quality-of-service (QoS) prediction. However, CF is vulnerable to shilling attackers who inject fake user profiles into the system. In this paper, we first present the shilling attack problem on CF-based QoS recommender systems for Web services. Then, a robust CF recommendation approach is proposed from a user similarity perspective to enhance the resistance of the recommender systems to the shilling attack. In the approach, the generally used similarity measures are analyzed, and the DegSim (the degree of similarities with top k neighbors) with those measures is selected for grouping and weighting the users. Then, the weights are used to calculate the service similarities/differences and predictions.We analyzed and evaluated our algorithms using WS-DREAM and Movielens datasets. The experimental results demonstrate that shilling attacks influence the prediction of QoS values, and our proposed features and algorithms achieve a higher degree of robustness against shilling attacks than the typical CF algorithms.     

融合领域知识的药店用户画像建模及应用

用户画像是对用户形象的勾勒与描述,现已广泛应用于睡眠会员唤醒,用户到店预测,个性化推荐等典型零售场景,药品不同于普通商品,包含较强的语义知识,现有用户画像主要从消费属性和静态属性出发,不能完全适用于药店销售和预测领域.本文提出了一种针对药品领域的用户画像模型UPP (persona of pharmacy user),在现有画像的基础上嵌入医药知识,利用规则,聚类,统计,实体识别等方法提取慢病、疾病、特殊病类、活动敏感度、用户价值、价格偏好等新标签.将所有标签融入一种基于聚类的群体划分方法,形成用户画像.实验表明,该模型相较于现有的用户画像模型,在消费行为预测场景下精准率提高了13%,更加适用于药店营销场景.     

Rating prediction for recommendation: Constructing user profiles and item characteristics using backpropagation

Recommender systems play an important role in quickly identifying and recommending most acceptable products to the users. The latent user factors and item characteristics determine the degree of user satisfaction on an item. While many of the methods in the literature have assumed that these factors are linear, there are some other methods that treat these factors as nonlinear; but they do it in a more implicit way. In this paper, we have investigated the effect of true nature (i.e., nonlinearity) of the user factors and item characteristics, and their complex layered relationship on rating prediction. We propose a new deep feedforward network that learns both the factors and their complex relationship concurrently. The aim of our study was to automate the construction of user profiles and item characteristics without using any demographic information and then use these constructed features to predict the degree of acceptability of an item to a user. We constructed the user and item factors by using separate learner weights at the lower layers, and modeled their complex relationship in the upper layers. The construction of the user profiles and the item characteristics, solely based on rating triples (i.e., user id, item id, rating), overcomes the requirement of explicit demographic information be given to the system. We have tested our model on three real world datasets: Jester, Movielens, and Yahoo music. Our model produces better rating predictions than some of the state-of-the-art methods which use demographic information. The root mean squared error incurred by our model on these datasets are 4.0873, 0.8110, and 0.9408 respectively. The errors are smaller than current best existing models’ errors in these datasets. The results show that our system can be integrated to any web store where development of hand engineered features for recommending products is less feasible due to huge traffics and also that there is a lack of demographic information about the users and the items.     

An Investigation into the Application of Different Performance Prediction Methods to Distributed Enterprise Applications

Response time predictions for workload on new server architectures can enhance Service Level Agreement–based resource management. This paper evaluates two performance prediction methods using a distributed enterprise application benchmark. The historical method makes predictions by extrapolating from previously gathered performance data, while the layered queuing method makes predictions by solving layered queuing networks. The methods are evaluated in terms of: the systems that can be modelled; the metrics that can be predicted; the ease with which the models can be created and the level of expertise required; the overheads of recalibrating a model; and the delay when evaluating a prediction. The paper also investigates how a prediction-enhanced resource management algorithm can be tuned so as to compensate for predictive inaccuracy and balance the costs of SLA violations and server usage.     

在线游戏用户的流失预测： 基于不平衡数据的采样方法比较和分析

流失用户预测问题在很多领域都是研究重点。目前主流的流失用户预测方法是使用分类法,即把用户是否会流失看作一个二分类问题来处理。该文提出了一个基于二分类问题解决的在线游戏流失用户预测方法。此方法除了总结了一些对在线游戏而言比较重要的可以用于流失预测的特征之外,也考虑到流失用户相对稀少的问题,在流失用户预测问题中引入了不平衡数据分类的思想。该文主要在流失预测中结合使用了基于采样法的不平衡数据处理策略,并对现有主要的几种采样算法进行了对比实验和分析。     

基于排队论的服务资源可用性相关研究综述

排队论在很多领域中解决了复杂的排队问题,文中首先对排队论的一般模型表示和常见模型进行了介绍.然后,简要对排队论解决的各类问题进行了归纳总结,重点对近年来有关排队论的服务资源可用性预测文献进行了综述,对排队模型在日常生活、云计算以及网络资源等场景中的应用进行总结.通过对文献进行综述找到服务与用户需求之间的关系,并对预测服...