Search

ObjectiveIn 2015 and 2016, the Canadian Journal of Emergency Medicine (CJEM) Social Media (SoMe) Team collaborated with established medical websites to promote CJEM articles using podcasts and infographics while tracking dissemination and readership.MethodsCJEM publications in the “Original Research” and “State of the Art” sections were selected by the SoMe Team for podcast and infographic promotion based on their perceived interest to emergency physicians. A control group was composed retrospectively of articles from the 2015 and 2016 issues with the highest Altmetric score that received standard Facebook and Twitter promotions. Studies on SoMe topics were excluded. Dissemination was quantified by January 1, 2017 Altmetric scores. Readership was measured by abstract and full-text views over a 3-month period. The number needed to view (NNV) was calculated by dividing abstract views by full-text views.ResultsTwenty-nine of 88 articles that met inclusion were included in the podcast (6), infographic (11), and control (12) groups. Descriptive statistics (mean, 95% confidence interval) were calculated for podcast (Altmetric: 61, 42-80; Abstract: 1795, 1135-2455; Full-text: 431, 0-1031), infographic (Altmetric: 31.5, 19-43; Abstract: 590, 361-819; Full-text: 65, 33-98), and control (Altmetric: 12, 8-15; Abstract: 257, 159-354; Full-Text: 73, 38-109) articles. The NNV was 4.2 for podcast, 9.0 for infographic, and 3.5 for control articles.DiscussionLimitations included selection bias, the influence of SoMe promotion on the Altmetric scores, and a lack of generalizability to other journals.ConclusionCollaboration with established SoMe websites using podcasts and infographics was associated with increased Altmetric scores and abstract views but not full-text article views.

The majority of data-driven fault detection approaches for automotive diagnostics are trained in a supervised manner. However, gathering diagnostics datasets to train supervised fault detection models is still challenging. Precise correlation of input data and defect events is non-trivial and requires extensive support of domain experts. Additionally, a strong imbalance caused by rare faulty events results in models which are biased towards non-faulty events. Hence, in this work, we propose a fully unsupervised data- driven diagnostics approach to detect faults in high frequency in-vehicle data. We transfer the concept of deep embedded clustering for static data to multivariate in- vehicle time series. We extend the approach by modifying the neural network architecture and comparing three similarity measures in the clustering layer, i.e., soft dynamic time warping, complexity invariant distance, and Euclidean distance. We further introduce an adapted target distribution to tackle imbalanced datasets. Our approach is evaluated on multivariate high frequency electronic control unit data of a test vehicle to detect pre-ignitions in high pressure turbocharged petrol engines. Current state-of-the-art time series clustering approaches are used as baselines for performance comparison. The results show that our approach is able to identify pre-ignitions without labels and outperforms the baselines by 10 percent in terms of accuracy.

It is highly expected that your car learns your individual preferences and context so that when it is cold, it automatically switches on the seat heating and when it is warm it switches on the air conditioning. This paper describes a simple computational method for learning individual preferences that has the advantage that it does not require to be trained in the vehicle on-board for every customer separately, while maintaining its ability to tune to every customer individually rather than defaulting to the average customer's preference. Our results show that our method outperforms the non-individual and clustering models, and demonstrates the feasibility of integration into the vehicle.

The automotive industry has been dominated by a few giant original equipment manufacturers, but these OEMs are being challenged by the software and service industries, including Apple and Google. Faced with such challenges from new players in the automobile segment, OEMs are strategically investing more in digital services and new business models, shifting focus toward software and service platforms, engaging in the user's daily life, and scalable cloud computing with machine learning and the Internet of Things (IoT). Here, the authors examine this shift toward digital services.

Vehicles have become essential tools of transport, offering a great opportunity to exploit the relationship between people and the car. This paper aims to solve an interesting problem, recognizing who the person is through their driving behaviors. Driver identification is useful for quite a few situations, such as car usage authentication, context-based recommendation, and determination of auto-insurance compensation. In this work, we propose WhozDriving, an approach that analyzes drivers’ driving behavior data and extract some sudden changes of driver behaviors as features which can be applied to distinguish different drivers. We propose a supervised learning method to detect anomaly driving trajectory from driving data. Experimental results on driving datasets show that our proposed approach is effective in terms of anomaly detection rate and misclassification anomaly rate.

We study the extent to which social ties between people can be inferred in large social network, in particular via active user interactions. In most online social networks, relationships are lack of meaning labels (e.g., "colleague" and "intimate friends") due to various reasons. Understanding the formation of different types of social relationships can provide us insights into the micro-level dynamics of the social network. In this work, we precisely define the problem of inferring social ties and propose a Partially-Labeled Pairwise Factor Graph Model (PLP-FGM) for learning to infer the type of social relationships. The model formalizes the problem of inferring social ties into a flexible semi-supervised framework. We test the model on three different genres of data sets and demonstrate its effectiveness. We further study how to leverage user interactions to help improve the inferring accuracy. Two active learning algorithms are proposed to actively select relationships to query users for their labels. Experimental results show that with only a few user corrections, the accuracy of inferring social ties can be significantly improved. Finally, to scale the model to handle real large networks, a distributed learning algorithm has been developed.

As a first step in the development of community trackers for large-scale online interaction, this paper shows how cohesive subgroup analysis using the Social Cohesion Analysis of Networks (SCAN; Chin and Chignell 2008) and Data-Intensive Socially Similar Evolving Community Tracker (DISSECT; Chin and Chignell 2010) methods can be applied to the problem of identifying cohesive subgroups and tracking them over time. Three case studies are reported, and the findings are used to evaluate how well the SCAN and DISSECT methods work for different types of data. In the largest of the case studies, variations in temporal cohesiveness are identified across a set of subgroups extracted from the inferred social network. Further modifications to the DISSECT methodology are suggested based on the results obtained. The paper concludes with recommendations concerning further research that would be beneficial in addressing the community tracking problem for online data.

The problem of identifying cohesive subgroups in social hypertext is reviewed. A computationally efficient three-step framework for identifying cohesive subgroups is proposed, referred to as the Social Cohesion Analysis of Networks (SCAN) method. In the first step of this method (Select), people within a social network are screened using a level of network centrality to select possible subgroup members. In the second step (Collect), the people selected in the first step are collected into subgroups identified at each point in time using hierarchical cluster analysis. In the third step (Choose), similarity modeling is used to choose cohesive subgroups based on the similarity of subgroups when compared across different points in time. The application of this SCAN method is then demonstrated in a case study where a subgroup is automatically extracted from a social network formed based on the online interactions of a group of about 150 people that occurred over a two-year period. In addition, this paper also demonstrates that similarity-based cohesion can provide a different, and in this case more compelling, subgroup representation than a method based on splitting a hierarchical clustering dendrogram using an optimality criterion.

Urban sensing and computing has become hot with the recent surge of Internet of Things, smart phones, and mobile social networks. Complementary to static sensing infrastructure, mobile crowd sensing (MCS) has become an important way to achieve large-scale urban sensing. However, few studies pay attention to public sensing to improve city management. In this paper, we propose an MCS-based framework to improve urban sensing and public facility management. We use a unique crowd sourced data which collects complaint calls from citizens in Xi'an city, as a case study for public facility sensing and management. We analyze the complaint call data from four different perspectives, and obtain several interesting findings. These findings are useful for applications such as public facility management, urban infrastructure maintenance and emergency sensing.

Participatory sensing campaigns encourage ordinary people to collect and share sensing data by using their portable smart devices like smartphones, iPad and iWatch. However, compared with static sensors that are deployed in the particular places by people that can provide high-quality data, the quality of sensing data that are collected from smart devices is uncontrollable, and it is costly to pay data usage without knowing the credibility of the collected data. In this paper, we propose to leverage the participant's reputation to maximize total data credibility under budget constraint for participatory sensing systems. Specifically, we formally define a maximum sensing data credibility problem and present a trustable participants selection algorithm to obtain a near optimal solution. Provided the real data set, it has been shown by simulation results that our proposed algorithm selects more credible sensing data than that of the other two comparison algorithms.

Recommender systems have changed the way people originally find products, information, and even their social circles. However, most existing research activities neglect its time-varying feature, i.e., the growing input data, the change of user behaviors. In order to sustain the high accuracy of recommendations, systems have to be updated regularly. However, the more often the update proceeds, the more cost of time and other computational resources. Thus, it is critical to strike the balance between accuracy and cost. In this paper, we propose an adaptive recommender system by using feedback control frameworks. The proposed solution continuously monitors its changes and estimates the loss of performance (in terms of accuracy) from two perspectives: data problem(data aging and data deficient) in training set, and changes of user behavior by “revisiting ratio”. When the benefit of performing an update exceeds the cost of resources, the system update itself. Theoretical analysis and extensive results by using a real data set are supplemented to show the advantages of the proposed system.

International audience; With the development of mobile sensing and mobile social networking techniques, mobile crowd sensing and computing (MCSC), which leverages heterogeneous crowdsourced data for large-scale sensing, has become a leading paradigm. Built on top of the participatory sensing vision, MCSC has two characteristic features: it leverages heterogeneous crowdsourced data from two data sources: participatory sensing and participatory social media; and it presents the fusion of human and machine intelligence in both the sensing and computing processes. This article characterizes the unique features and challenges of MCSC. We further present early efforts on MCSC to demonstrate the benefits of aggregating heterogeneous crowdsourced data

Location-based services has attracted attentions from both industry and academia. The development of position tracking technologies and the increasing popularity of smart phones has collected large amounts of contexts. An important issue is how to leverage the rich contexts to predict a user's need accurately. In this paper, we propose a novel approach to predict the product type that a user will first click in an e-commerce application, after they update their location manually. Our proposed approach models the problem as a multi-label classification. We introduce three sets of features including location feature, time feature and behavioral feature. We use the Periodica algorithm [10], which was designed to mine the periodic behaviors of moving objects, to generate a series of periodicity templates. The templates are further exploited as behavioral features. Finally, we design several experiments using a real world dataset collected by an e-commerce application called WuXianGouXiang, which is developed by Nokia Research Center, Beijing and was released in September 2011. We have obtained a dataset from the service logs and the dataset contains over 3000 registered shops and 20000 users. Our experimental results demonstrate that the three sets of features contribute significantly to the classification of different users and the best result is achieved when using all of them.

With mobile devices and wireless technology becoming pervasive at conferences, conference attendees can use location-based mobile social networking applications such as Foursquare and Gowalla to share their location and content with others. How to use the mobile devices and the indoor positioning technology to help the conference participants enhance the real-world interactions of people and improve efficiency during the conference? In this paper, we report our work in Nokia Find & Connect (NF&C) to solve this problem where we use location and encounters, together with the conference basic services, all through a mobile UI. To demonstrate the usefulness of the NF&C system, we conducted a field trial at the 7th International Conference on Ubiquitous Intelligence and Computing and Autonomic and Trusted Computing (UIC/ATC 2010). Results show that NF&C can help the conference participants enhance the real-world interactions of people and improve efficiency at a conference effectively.

With growing popularity of social networks, people’s locations are being used for providing rich mobile social services, mostly with outdoor positioning using GPS. We present a mobile social service deployed in an indoor office environment called “Nokia Find & Connect” (NF&C). In NF&C, employees’ locations are tracked using Wi-Fi to efficiently find, reserve and manage office resources like meeting rooms and desks and to easily connect other colleagues through scheduled interactions like having a meeting together and/or unscheduled yet implicit interactions like ephemeral encounters between two people. We present the user interface and functional architecture of NF&C and discuss a user study that we conducted in our office.

This chapter examines the problem of tracking community in social networks inferred from online interactions by tracking evolution of known subgroups over time. Finding subgroups within social networks is important for understanding and possibly influencing the formation and evolution of online communities. A variety of approaches have been suggested to address this problem and the corresponding research literature on centrality, clustering, and optimization methods for finding subgroupings is reviewed. This review will include a critical analysis of the limitations of past approaches. The focus of the chapter will then turn to novel methods for tracking online community interaction. First, the method proposed by Chin and Chignell called SCAN will be briefly introduced, where a combination of heuristic methods is used to identify subgroups in a manner that can potentially scale up to very large social networks. Then, we present the DISSECT method where multiple known subgroups within a social network are tracked in terms of similarity-based cohesiveness over time. The DISSECT method relies on cluster analysis of snapshots of network activity at different points in time followed by similarity analysis of subgroup evolution over successive time periods. The DISSECT method can be supplemented with behavioral measures of sense of community where administration of a questionnaire is feasible. Finally, we conclude the chapter with a discussion on possible applications and use of the DISSECT method.

This paper describes a business user study using a packaged suite of mobile Web services and applications deployed at a real-time event, the Beijing Olympic Games in August 2008. These applications were an Olympics guide, menu reader, phrasebook, Sports Tracker [3], photo sharing on Ovi [4], and Nokia Maps [2]. To evaluate its feasibility and use, we used logging, surveys, and statistical analysis for collecting and analyzing the data. We discovered that guests found the Olympics guide application to be the most popular, followed by Nokia Maps and then photo sharing on Ovi. The results demonstrate that the techniques used in our evaluation can be used to determine the type of applications that are relevant to consumers at a real-time event, and suggests ways for improving the mobile application design and user experience.

Mobile devices are increasingly being used to access Web content but lack the resources for proper presentation to the user. To address this problem, content is typically adapted to be more suitable for a mobile environment. Community-driven adaptation (CDA) is a novel approach to automatic content adaptation for mobile devices that adapts content based on feedback from users. CDA groups users into communities based on common characteristics, and assumes that users of the same community have similar adaptation requirements. CDA learns how to adapt content by observing how members of a community alter adapted content to make it more useful to them. Experiments that consider the idealized case, where all users perform the same task, show that CDA can reduce wastage of network bandwidth by up to 90% and requires less user interaction to correct bad adaptation decisions compared with existing approaches to automatic content adaptation.

This paper proposes an architectural framework for integrating services within an enterprise and accessing them from mobile devices in a pervasive-computing environment. Present network environments are prone to failures from disconnections and device crashes. Being able to maintain service transaction while moving to different locations and wireless networks (known as service roaming), is currently a major objective in pervasive and mobile computing research. This paper presents a framework (called m-Roam) that allows for mobile clients to perform service invocation and roaming. The framework addresses this problem by introducing a proxy-based architecture so that invocation context and transaction state can be maintained or updated accordingly, while the user roams in different locations.

Social network services (SNSs) are developing at an explosive speed, which makes it easy for people to be closely connected. As a result, a new SNS type, ephemeral social network (ESN), is emerging to capture the ephemeral interactions and meetings that occur in environments such as conferences and workplaces. Most ESNs require the positioning for capturing the physical proximity between people, which impacts much the scalability, privacy protection, and the cost of the system. In this paper, we propose a scalable location-sensing model based on RFID-sensing architecture for ESN in consideration of four aspects of requirements, that is, the usability, QoS, scalability, and privacy. The model includes the perspectives of the privacy, architecture, deployment, and positioning algorithms, which can meet the four key requirements. A prototyping conference-type ESN system was also developed in this paper and was tested in the ACM UbiComp 2011 conference, which verified that the model works well with good scalability, low cost, and customized privacy protection.

The ACM Hypertext Conference is the main venue for high quality, peer-reviewed, double-blind research on "linking" and interconnectivity. The Web, the Semantic Web, Web 2.0, Social Networks, and Web Science all demonstrate the value of the link concept.