83,543,095 research outputs found
Spatio-Cohesive Service Selection Using Machine Learning in Dynamic IoT Environments
No full textIn an Internet of Things (IoT) environment, a user may need to receive a set of services from the environment to accomplish his or her own task. Then, appropriate IoT devices that are necessary to provide the services should be located and accessed in a way that the functional effects of the IoT services can be delivered effectively to the user. Especially, it is more challenging to ensure the successful delivery of service effects when users and/or IoT devices have mobility in the environment. In this paper, we extend our previous work of spatially-cohesive service selection by adopting a reinforcement learning technique to effectively deal with highly dynamic situations in mobile IoT environments. We propose a reinforcement learning agent that selects services in terms of associating IoT devices in a dynamic manner while optimizing metrics such as spatio-cohesiveness and number of hand-overs during the period of providing the services. Our approach is evaluated by simulating the agent, and the results show that the agent successfully learns the optimal policy of the service selection
Service consumption planning for efficient service migration in mobile edge computing environments
No full textMobile Edge Computing (MEC) is the concept of placing a cloud computing server to run at the network edge near the user. The faster the user moves, the higher the probability of leaving the coverage of an MEC server. Seamless service migration should be implemented to assure service continuity. However, the time required for service migration is not negligible; thus, the frequent service migration should be avoided. In this study, we define the service consumption plan optimization problem and prove its NP-completeness. To solve the problem, we proposed a genetic algorithm-based method and conducted vast experiments to evaluate the algorithm with respect to other baselines based on real-world data sources. The results showed that the proposed method generates improved service consumption plans than other baselines in all the scenarios we set out. © 2021 ACM
Mashup Recommendation for Trigger Action Programming
No full textIf This Then That (IFTTT) is a popular platform that deploys mashed-up applications for end users using trigger-action programming (TAP) paradigm. To date, there are about 135 thousand mashup creators who have shared recipes for developing applications using TAP, and around 24 million mashups have been adopted by users. Up to this date, research has not focused on recommending personalized mashups for the users. In this work, we propose a model for mashup recommendation for Trigger Action Programming. We tested our recommendation algorithm using the 200,000 recipes dataset from the IFTTT platform and compared its performance with other popular algorithms for content recommendation
Task-oriented Approach to Guide Visually Impaired People During Smart Device Usage
No full textVisually impaired persons often experience limited user accessibility regarding smart devices. Particularly, it is challenging for such persons to precisely touch a button or icon to operate smart devices. Existing studies have reported a lack of step-by-step guidance for visually impaired users to perform tasks on smart devices. Although prototypes and non-visual interface systems were implemented to aid blind people in using mobile phones, a step-by-step guidance system was not developed. Moreover, devices other than smartphones were not emphasized. To address these limitations, we propose a novel guidance system that employs an object detection method for recognizing the control panels of smart devices and offers appropriate guidelines based on the user's targeted task. A user study conducted with 20 participants indicated that users were more satisfied with our guidance system than with existing voice assistants
A Conversational Approach for Modifying Service Mashup in IoT Environments
No full textExisting conversational approaches for Internet of Things (IoT) service mashup do not support modification because of the usability challenge, although it is common for users to modify the service mashups in IoT environments. To support the modification of IoT service mashups through conversational interfaces in a usable manner, we propose the conversational mashup modification agent (CoMMA). Users can modify IoT service mashups using CoMMA through natural language conversations. CoMMA has a two-step mashup modification interaction, an implicature-based localization step, and a modification step with a disambiguation strategy. The localization step allows users to easily search for a mashup by vocalizing their expressions in the environment. The modification step supports users to modify mashups by speaking simple modification commands. We conducted a user study and the results show that CoMMA is as effective as visual approaches in terms of task completion time and perceived task workload for modifying IoT service mashups
Multi-criteria matrix localization and integration for personalized collaborative filtering in IoT environments
No full textCollaborative filtering (CF)-based recommender systems can be used to deal with the complexity problem of users when they want to identify possible tasks on the fly and perform desired tasks by using various smart objects in Internet of Things (IoT) environments. However, in order to use CF-based recommender systems, users need to provide their feedbacks and there are usually more than one criterion considered when users choose an item. Although there have been studies of multi-criteria recommendations, existing approaches require multi-criteria ratings that are explicitly given by users. It is usually a burden for a user to provide more than one instance of feedback on an item; therefore, user feedback datasets are usually sparse when users are asked to provide multi-criteria ratings. Due to the sparsity of multi-criteria rating data, the similarity measurements used by the existing approaches may produce biased results, possibly leading to degradation of the recommendation accuracy. This problem becomes worse as the sparsity of a dataset increases. To alleviate the effects of the data-sparsity problem, and to take advantage of using multi-criteria ratings, we proposed a multi-criteria matrix localization and integration (MCMLI) approach for collaborative filtering in this paper. The main goal of MCMLI is to find cohesive user-item subgroups (CUISs) for each criterion from sparse data, and to predict users' interests for each criterion in a more precise manner. The proposed approach is composed of three phases. At the first phase, a given user-item matrix is divided into a set of CUIS matrices, each of which is organized with correlated users and items for each criterion. MCMLI repeats this CUIS generation process until the generated subgroups cover all elements of the given user-item matrix. To generate prediction results for each criterion, MCMLI then predicts user ratings on new items for each CUIS and aggregates the prediction results to make recommendations to users. To enable personalized recommendations, during the aggregation process, each user's preferences on multiple criteria are weighted differently according to the number of CUISs to which the user belongs. We demonstrate the effectiveness of our approach by conducting an experiment with real-world datasets from TripAdvisor and Yahoo! Movies. The experimental results show that MCMLI outperforms existing multi-criteria collaborative-filtering-based recommendation methods in terms of the recommendation accuracy. In addition, unlike the existing multi-criteria recommendation approaches, even when the sparsity level of a dataset increases, the recommendation accuracy of MCMLI does not decrease significantly.
Urban Event Detection from Spatio-temporal IoT Sensor Data Using Graph-Based Machine Learning
No full textThe accurate detection and handling of urban events such as emergency incidents are critical to improve the safety and convenience of people's life in urban environments. Recently, it has become possible to detect and handle urban events in an effective manner by analyzing the detailed urban data collected from Internet of Things (IoT) sensors. Especially, some recent works investigated the use of spatio-temporal sensor data for urban event detection. However, we found there is one challenge of having less accuracy of detecting urban events as the granularity of processing data in the spatial dimension becomes finer. To meet the challenge, we propose a novel graph-based approach that analyzes geo-spatial characteristics of urban sensor data over time to keep the accuracy of detecting urban anomaly in finer-grained geo-spatial scales, by identifying and exploiting regions that have abnormal urban dynamics. Through experiments using real-world urban datasets, we show our approach effectively addresses the challenge and outperforms the popular machine-learning-based urban event detection methods
Auto-labeling of sensor data using social media messages
No full textRecently, the deployment of various Internet of Things (IoT) sensors has encouraged smart cities to accumulate a large volume of data. When machine learning models utilize such accumulated raw data to predict events and situations, various systems of a smart city, such as traffic accident management systems, can be further developed by utilizing the predicted events and situations. However, although there has been a large volume of various IoT raw data on smart cities, such data do not have labels that are related to events or situations. Data with meaningful labels are required for the training of the models. Because these sensor data do not have meaningful labels, the data cannot be utilized into the models. There are several existing methods for labeling, but they have different drawbacks. In this study, we investigate the feasibility of utilizing social media messages to extract meaningful labels for machine learning to predict events and situations in smart city environments. As a case study, we compared the extracted labels from social media messages with the events and situations found in announced traffic news, and other articles. The results show that it is feasible to utilize social media messages as a source for meaningful labels of events and situations. We also propose an improved clustering algorithm using an outlier detection technique to extract meaningful labels in a more robust way. Furthermore, for other researchers who want to utilize the IoT raw data, we analyze and release the refined sensor data on which there were unknown noise. © 2021 ACM
An Empirical Study of Utility and Disclosure Risk for Tabular Data Synthesis Models: In-Depth Analysis and Interesting Findings
No full textThe ever-growing accumulation of data in various applications has spurred research into privacy-enhancing technologies. Synthetic data, in particular, has gained significant attention for enhancing machine learning model performance while preserving personal information. Although synthetic data studies have been on the rise, there are no clear criteria for how to measure the utility and disclosure risk of synthetic data. Furthermore, although many existing studies have primarily concentrated on image data synthesis models, there's a notable scarcity of research on tabular data synthesis models, particularly concerning disclosure risk. This is crucial in domains such as finance, which heavily rely on tabular datasets containing sensitive information. In this paper, we perform in-depth analysis of utility and disclosure risk index from classical to state-of-the-art tabular data synthesis models in terms of different metrics and various types of datasets. Our interesting findings can be summarized as follows: (1) Synthetic data's utility tends to increase as the proportion of continuous attributes in the original data decreases, (2) Conversely, disclosure risk rises with a lower proportion of continuous attributes in the original data, (3) As the volume of synthetic data grows, both utility and disclosure risk metrics generally increase, (4) An inverse relationship is observed between the sparsity of original data and a specific utility metric, and (5) Notably, we discover that Targeted Correct Attribution Probability (TCAP), a widely-used disclosure risk metric, fails to measure certain outlier records that are potential vulnerabilities for malicious attacks
- …
