Historical Usage Data Research Articles

A Newsvendor Approach to Design of Surgical Preference Cards

Surgical procedures require a large number of consumable supplies that need to be kept in hospital inventory and transported to the operating rooms (OR) before the surgery. A surgical preference card (SPC) provides a list of items to be prepared for each surgery. For each item, a SPC also specifies how many should be taken to the OR (fill quantity). As the usage of most consumables in the OR is subject to uncertainty, the cards also specify how many of the filled items should be opened at the beginning of the surgery (open quantity). The fill and open quantities control the flow of consumables between the hospital inventory and the ORs and directly affect the wastage in ORs. In this work, we formulate the problem of determining the fill and open quantities on the preference cards as a stochastic optimization problem, where the objective is to minimize a weighted sum of the expected wastage and operational costs. We show that, as in the newsvendor problem, the optimal solution for the fill and open quantities takes the form of critical quantiles of the item usage distribution in the OR. The solution form together with historical usage data provide a data-driven approach to design of SPCs, as well as insights on the value of including an open decision. We demonstrate our approach using extensive numerical experiments and real usage data from a Canadian hospital. The results suggest a potential for significant reduction of wastage and operational costs in the ORs.

Recommender systems for mental health apps: advantages and ethical challenges.

Recommender systems assist users in receiving preferred or relevant services and information. Using such technology could be instrumental in addressing the lack of relevance digital mental health apps have to the user, a leading cause of low engagement. However, the use of recommender systems for digital mental health apps, particularly those driven by personal data and artificial intelligence, presents a range of ethical considerations. This paper focuses on considerations particular to the juncture of recommender systems and digital mental health technologies. While separate bodies of work have focused on these two areas, to our knowledge, the intersection presented in this paper has not yet been examined. This paper identifies and discusses a set of advantages and ethical concerns related to incorporating recommender systems into the digital mental health (DMH) ecosystem. Advantages of incorporating recommender systems into DMH apps are identified as (1) a reduction in choice overload, (2) improvement to the digital therapeutic alliance, and (3) increased access to personal data & self-management. Ethical challenges identified are (1) lack of explainability, (2) complexities pertaining to the privacy/personalization trade-off and recommendation quality, and (3) the control of app usage history data. These novel considerations will provide a greater understanding of how DMH apps can effectively and ethically implement recommender systems.

Exploiting Group Information for Personalized Recommendation with Graph Neural Networks

Personalized recommendation has become more and more important for users to quickly find relevant items. The key issue of the recommender system is how to model user preferences. Previous work mostly employed user historical data to learn users’ preferences, but faced with the data sparsity problem. The prevalence of online social networks promotes increasing online discussion groups, and users in the same group often have similar interests and preferences. Therefore, it is necessary to integrate group information for personalized recommendation. The existing work on group-information-enhanced recommender systems mainly relies on the item information related to the group, which is not expressive enough to capture the complicated preference dependency relationships between group users and the target user. In this article, we solve the problem with the graph neural networks. Specifically, the relationship between users and items, the item preferences of groups, and the groups that users participate in are constructed as bipartite graphs, respectively, and the user preferences for items are learned end to end through the graph neural network. The experimental results on the Last.fm and Douban Movie datasets show that considering group preferences can improve the recommendation performance and demonstrate the superiority on sparse users compared

Efficient on-the-fly Web bot detection

A large fraction of traffic on present-day Web servers is generated by bots — intelligent agents able to traverse the Web and execute various advanced tasks. Since bots’ activity may raise concerns about server security and performance, many studies have investigated traffic features discriminating bots from human visitors and developed methods for automated traffic classification. Very few previous works, however, aim at identifying bots on-the-fly, trying to classify active sessions as early as possible. This paper proposes a novel method for binary classification of streams of Web server requests in order to label each active session as “bot” or “human”. A machine learning approach has been developed to discover traffic patterns from historical usage data. The model, built on a neural network, is used to classify each incoming HTTP request and a sequential probabilistic analysis approach is then applied to capture relationships between subsequent HTTP requests in an ongoing session to assess the likelihood of the session being generated by a bot or a human, as soon as possible. A performance evaluation study with real server traffic data confirmed the effectiveness of the proposed classifier in discriminating bots from humans at early stages of their visits, leaving very few of them undecided, with very low number of false positives.

Development of Personalized Learning Resources Recommendation System Based on Knowledge Graph

This study focuses on how to efficiently and accurately recommend personalized learning resources for users when they are in the face of massive learning resources. A recommendation system is developed with software engineering methods. Knowledge graph technology is integrated in the system; curriculum knowledge graphs are constructed to solve the problems of semi-structured data storage and knowledge fragmentation. Two kinds of recommendation methods are adopted to achieve the goal of learners' personalized learning, one is Euclidean distance recommendation algorithm based on user behavior graph library, the other is learning mode recommendation algorithm and sequential mode recommendation algorithm based on user session library. The recommendation system maintains the interpret-ability and the accuracy of recommendation based on user historical behavior data; and realizes recommendation based on user session library in the context of lacking users’ historical data.

Improved Algorithm of Similarity Measure Based on Matrix Factorization Filling and Filling Confidence

Although collaborative filtering recommendation has been very mature in practical application, the problem of data sparsity has been difficult to solve. Sparse data seriously weakens the accuracy of similarity measurement, and also affects the recommendation accuracy of recommendation system. Therefore, this paper makes a deep study on this problem and proposes an improved similarity measure algorithm based on matrix factorization filling and filling confidence. The core idea of the algorithm is to fill the original sparse data set first, and then calculate the similarity based on the filling. In the process of matrix filling, considering that the matrix decomposition model can effectively alleviate the sparsity of data by associating users and items with hidden features, this paper chooses the BiasSVD matrix decomposition model as the basis for filling improvement. Considering that the value of user's interest or behavior will decline with the change of time, the user's recent interaction data can better reflect their current interest than the long-term interaction data, so the improved algorithm integrates the time decline function into the BiasSVD algorithm when filling, and uses the time decline function to give different weights to the user's historical data at different stages, In order to achieve better filling effect. At the same time, after filling, considering that the traditional matrix filling algorithm ignores the credibility difference between the real data and the filled data, and there is no distinction in the next step of similarity calculation, which affects the performance of the recommendation. Therefore, the concept of filling confidence is introduced in the calculation of similarity, which fully distinguishes the reference of the real score and the predicted score.

Diversified service recommendation with high accuracy and efficiency

Collaborative filtering-based recommender systems are regarded as an important tool to predict the items that users will appreciate based on the historical usage of users. However, traditional recommendation solutions often pay more attentions to the accuracy of the recommended items while neglect the diversity of the final recommended list, which may produce partial redundant items in the recommended list and as a result, decrease the satisfaction degree of users. Moreover, historical usage data for recommendation decision-makingoften update frequently, which may lead to low recommendation efficiency as well as scalability especially in the big data environment. Considering these drawbacks, a novel method called DivRec_LSH is proposed in this paper to achieve diversified and efficient recommendations, which is based on the historical usage records and the Locality-Sensitive Hashing (LSH) technique. Finally, we compare our method with existing methods on the MovieLens dataset. Experiment results indicate that our proposal is feasible in addressing the triple dilemmas of recommender systems simultaneously, i.e., high efficiency, accuracy and diversity.

Living with I-Fabric: Smart Living Powered by Intelligent Fabric and Deep Analytics

With the emergence of next-generation fiber materials and the development of sensors and wearable technologies, new breakthroughs also emerge in the field of wearable fabrics. By combining wearable fabrics with sensors, artificial intelligence and other technologies, the new intelligent fabrics create an intelligent, flexible and adaptive performance, thus meeting people's needs better and becoming more acceptable to users than traditional fabrics. Nevertheless, present intelligent fabrics have some shortcomings, such as insufficient interactive ability, the narrow scope of applications, and poor ability in synchronizing physical and psychological health data. Therefore, in this article, we propose the Living with I-Fabric system, which represents smart living powered by intelligent fabric and deep analytics. The proposed system is capable of solving problems that are beyond the scope of existing fabrics and further improve the user's experience. Combining multiple smart accessories and cognitive devices, together with smartphones and multi-level clouds, the proposed system builds a comprehensive and sustainable health monitoring and guidance ecosystem. Through AI algorithms on the edge and remote clouds, the user's historical data is analyzed in depth, and additional valued function services can be provided for users.

Latent Variable Models for Integrated Analysis of Credit and Point Usage History Data on Rewards Credit Card System

Recently, credit cards with point rewards functions (rewards credit cards) are widely used. Credit card companies can collect the users’ usage log data of various stores in multiple industries. The purposes of possessing a credit card varies depending on each user such as to use only the credit function, to use both the credit and point rewards functions, etc. Moreover, credit cards can be used in various situations in users’ lives, and the purchase history of each user is diverse. Focusing on the diversity of both card possessing purposes and purchasing behavior for each user, we propose two latent class models representing these diversities in this research.

Tree-Based Real-Time Advertisement Recommendation System in Online Broadcasting

The viewing time of media content per week through TV is still dominant. Users are exposed to numerous advertisements, such as commercials, electronic home shopping, product placement (PPL), and T-Commerce while watching TV programs. Most of the advertisement systems provide a good overview of products. However, traditional advertising services do not consider user preferences, meaning it is difficult to expect anything more than mere exposure to them. We can adopt a recommendation system to predict the preference. However, existing recommendation systems find it difficult to satisfy the real-time requirements of online broadcasting because of the large overhead incurred in preference prediction processes. In this paper, we propose a real-time recommendation system to provide personalized advertisements. The proposed system generates tree models based on user historical data. To reduce the overhead of preference prediction, we introduce a sorted HashMap that enables fast tree searches. For sophisticated preference prediction, the proposed system normalizes the users' preferences by considering the characteristics of their tree model. Finally, we conduct experiments to evaluate the performance of the proposed tree-based recommendation system.

DLTSR: A Deep Learning Framework for Recommendations of Long-Tail Web Services

With the growing popularity of web services, more and more developers are composing multiple services into mashups. Developers show an increasing interest in non-popular services (i.e., long-tail ones), however, there are very scarce studies trying to address the long-tail web service recommendation problem. The major challenges for recommending long-tail services accurately include severe sparsity of historical usage data and unsatisfactory quality of description content. In this paper, we propose to build a deep learning framework to address these challenges and perform accurate long-tail recommendations. To tackle the problem of unsatisfactory quality of description content, we use stacked denoising autoencoders (SDAE) to perform feature extraction. Additionally, we impose the usage records in hot services as a regularization of the encoding output of SDAE, to provide feedback to content extraction. To address the sparsity of historical usage data, we learn the patterns of developers’ preference instead of modeling individual services. Our experimental results on a real-world dataset demonstrate that, with such joint autoencoder based feature representation and content-usage learning framework, the proposed algorithm outperforms the state-of-the-art baselines significantly.

Improved LSH for privacy-aware and robust recommender system with sparse data in edge environment

Recommender systems have become a popular and effective way to quickly discover new service items that are probably preferred by prospective users. Through analyzing the historical service usage data produced in the past, a recommender system can infer the potential user preferences and make accurate recommendations accordingly. However, in the edge environment, the service usage data stored in each edge server are often very sparse, which may result in expected cold-start problems. Besides, in the edge environment, the data required to make an optimal service recommendation decision are often stored in different edge clients or servers, which require additional privacy-preservation strategies to secure the sensitive data involved. Considering the above two drawbacks, traditional locality-sensitive hashing (LSH) is improved to be multi-probing LSH and then introduced to aid the recommendation process so as to guarantee the security and robustness of recommender systems. Experiments conducted on well-known dataset prove the effectiveness and efficiency of the work.

Online Diagnosis of Performance Variation in HPC Systems Using Machine Learning

As the size and complexity of high performance computing (HPC) systems grow in line with advancements in hardware and software technology, HPC systems increasingly suffer from performance variations due to shared resource contention as well as software- and hardware-related problems. Such performance variations can lead to failures and inefficiencies, which impact the cost and resilience of HPC systems. To minimize the impact of performance variations, one must quickly and accurately detect and diagnose the anomalies that cause the variations and take mitigating actions. However, it is difficult to identify anomalies based on the voluminous, high-dimensional, and noisy data collected by system monitoring infrastructures. This paper presents a novel machine learning based framework to automatically diagnose performance anomalies at runtime. Our framework leverages historical resource usage data to extract signatures of previously-observed anomalies. We first convert collected time series data into easy-to-compute statistical features. We then identify the features that are required to detect anomalies, and extract the signatures of these anomalies. At runtime, we use these signatures to diagnose anomalies with negligible overhead. We evaluate our framework using experiments on a real-world HPC supercomputer and demonstrate that our approach successfully identifies 98 percent of injected anomalies and consistently outperforms existing anomaly diagnosis techniques.

A Two-Stage Destination Prediction Framework of Shared Bicycles Based on Geographical Position Recommendation

In this study, we investigate the destination prediction problem for the trips of shared bicycles, which can assist authorities in overcoming the existing challenges like indiscriminate parking, bicycle rebalancing, etc. An integral part of E-business, the recommendation system can predict the buyers' needs and recommend personalized contents for each buyer. Likewise, prediction of travelers' destinations is essentially a geographical position recommendation. In this paper, we present a two-stage destination prediction framework inspired by the recommender system. It is an innovative framework for the destination prediction problem in transportation studies. The architecture is highly flexible and extendable while it utilizes the user's historical data such as departure time, geographic position, and traveling behavior. The experiments demonstrate that the proposed method performs well in the real situation.

E-Memory Choice Architecture

Twitter archiving systems have been developed to preserve users' tweets. The available methods of organizing tweets for curation include the hashtag, user ID, and keywords. These can be viewed as memory encoding symbols supporting future retrieval of users' social media memories. As Twitter has become a global social media platform, online Twitter archiving systems have transformed from an open platform for archiving tweets to an integrated service managing multiple accounts across platforms. With the changing business models of Twitter archiving systems, usage data has become unavailable publicly. This study collected historical usage data from the API of an online Twitter archiving system, TwapperKeeper, before its acquisition by Hootsuite in September 2011. The valuable system usage data allowed this study to examine the tweet archiving preferences of early Twitter adopters. By mapping adoption-diffusion and use-diffusion models into the web information architecture of the online archiving system, this study analyzed user choice architecture through the system function use.

Network Bandwidth Utilization Prediction Based on Observed SNMP Data

Bandwidth requirement prediction is an important part of network design and service planning. The natural way of predicting bandwidth requirement for existing network is to analyze the past trends and apply appropriate mathematical model to predict for the future. For this research, the historical usage data of FWDR network nodes of Nepal Telecom is subject to univariate linear time series ARIMA model after logit transformation to predict future bandwidth requirement. The predicted data is compared to the real data obtained from the same network and the predicted data has been found to be within 10% MAPE. This model reduces the MAPE by 11.71% and 15.42% respectively as compared to the non-logit transformed ARIMA model at 99% CI. The results imply that the logit transformed ARIMA model has better performance compared to non-logit-transformed ARIMA model. For more accurate and longer term predictions, larger dataset can be taken along with season adjustments and consideration of long term variations.Journal of the Institute of Engineering, 2017, 13(1): 160-168

A two-stage locality-sensitive hashing based approach for privacy-preserving mobile service recommendation in cross-platform edge environment

With the increasing popularity of service computing paradigm, tremendous resources or services are emerging rapidly on the Web, imposing heavy burdens on the service selection decisions of users. In this situation, recommendation (e.g., collaborative filtering) has been considered as one of the most effective ways to alleviate such burdens. However, in the mobile and edge environment, the service recommendation bases, i.e., historical service usage data are often generated from various mobile devices (e.g., Smartphone and PDA) and stored in different edge platforms. Therefore, effective collaboration between these distributed edge platforms plays an important role in the successful mobile service recommendation. Such a cross-platform collaboration process often faces the following two challenges. First, a platform is often reluctant to release its data to other platforms due to privacy concerns. Second, the collaboration efficiency is often low when the data in each platform update frequently. In view of these two challenges, we introduce MinHash, an instance of Locality-Sensitive Hashing (LSH), into service recommendation, and further put forward a novel privacy-preserving and scalable mobile service recommendation approach based on two-stage LSH, namedSerRectwo-LSH. Finally, extensive experiments are conducted on WS-DREAM, a real distributed service quality dataset, and the evaluation results demonstrate that both the service recommendation accuracy and the scalability have been significantly improved while privacy preservation is guaranteed.

Dataset Popularity Prediction for Caching of CMS Big Data

The Compact Muon Solenoid (CMS) experiment at the European Organization for Nuclear Research (CERN) deploys its data collections, simulation and analysis activities on a distributed computing infrastructure involving more than 70 sites worldwide. The historical usage data recorded by this large infrastructure is a rich source of information for system tuning and capacity planning. In this paper we investigate how to leverage machine learning on this huge amount of data in order to discover patterns and correlations useful to enhance the overall efficiency of the distributed infrastructure in terms of CPU utilization and task completion time. In particular we propose a scalable pipeline of components built on top of the Spark engine for large-scale data processing, whose goal is collecting from different sites the dataset access logs, organizing them into weekly snapshots, and training, on these snapshots, predictive models able to forecast which datasets will become popular over time. The high accuracy achieved indicates the ability of the learned model to correctly separate popular datasets from unpopular ones. Dataset popularity predictions are then exploited within a novel data caching policy, called PPC (Popularity Prediction Caching). We evaluate the performance of PPC against popular caching policy baselines like LRU (Least Recently Used). The experiments conducted on large traces of real dataset accesses show that PPC outperforms LRU reducing the number of cache misses up to 20% in some sites.

Improving Cluster-based Methods for Usage Anticipation by the Application of Data Transformations

The wide adoption of Internet of Things (IoT) infrastructure in recent years has allowed capturing data from systems that make intensive use of electrical power or consumables typically aiming to create predictive models to anticipate a system’s demand and to optimize system control, assuring the service while minimizing the overall consumption. Several methods have been presented to perform usage anticipation; one promising approach involves a two step procedure: profiling, which discovers typical usage profiles; and, prediction that detects the most likely profile given the current information. However, depending on the problem at hand, the number of observations to characterize a profile can increase greatly, causing high dimensionality, thus complicating the profiling step as the amount of noise and correlated features increase. In addition, the profile detection uncertainty increases, as the cluster intra-variability becomes larger and the distances between the centroids become similar. To overcome the difficulties that a usage profile with high dimensionality poses, we developed a methodology that finds the intrinsic dimensionality of a dataset, containing binary historical usage data, by performing dimensionality reductions to improve the profiling step. Then, the profile detection step makes use of the transformed actual data to accurately detect the current profile. This paper describes the implementation details of the application of such techniques by the analysis of two use cases: (1) usage prediction of a laser cutter machine; and, (2) occupancy prediction in an office environment. We observed that the dataset dimensionality and the cluster intra-variability was greatly reduced, making the profile detection less prone to errors. In conclusion, the implementation of methodologies to enhance the separability of the original data by dimensionality transformations improves the profile discovery and the subsequent actual profile detection.

Cognitively Adjusting Imprecise User Preferences for Service Selection

Most state-of-the-art service selection approaches assume user preferences can be provided by the target user with sufficient precision and ignore historical service usage data for all users. It is desirable for ordinary users to possess a new service selection approach that can recommend satisfactory services to them even when their service selection preferences are specified imprecisely in terms of vagueness, inaccuracy, and incompleteness. This paper proposes a novel service selection approach that resolves the imprecise characteristics of user preferences and can recommend satisfactory services for users with varying cognitive levels in terms of service experience. The proposed service selection approach is comprised of four major tasks: 1) employ user-friendly linguistic variables to collect apparent user preferences (AUP) and convert the linguistic variables to standardized fuzzy weights as AUP weights; 2) evaluate all users’ respective cognitive levels for the target service type and obtain the cognitive level threshold for that type of services; 3) adjust the AUP weights based on the calculated cognitive levels and the threshold, and supplement the potential user preferences weights; and 4) prioritize candidate services per a user satisfaction maximization objective. In-depth comparative experimental evaluations were performed using two real-world datasets. The results show that our service selection model outperforms three other representative ones and could provide a stable and reliable selection of services for the users with low service cognitive levels.