Discovering user communities on the Internet using unsupervised machine learning techniques

Interest in the analysis of user behaviour on the Internet has been increasing rapidly, especially since the advent of electronic commerce. In this context, we argue here for the usefulness of constructing communities of users with common behaviour, making use of machine learning techniques. In particular, we assume that the users of any service on the Internet constitute a large community and we aim to construct smaller communities of users with common characteristics. The paper presents the results of three case studies for three different types of Internet service: a digital library, an information broker and a Web site. Particular attention is paid on the different types of information access involved in the three case studies: query-based information retrieval, profile-based information filtering and Web-site navigation. Each type of access imposes different constraints on the representation of the learning task. Two different unsupervised learning methods are evaluated: conceptual clustering and cluster mining. One of our main concerns is the construction of meaningful communities that can be used for improving information access on the Internet. Analysis of the results in the three case studies brings to surface some of the important properties of the task, suggesting the feasibility of a common methodology for the three different types of information access on the Internet.     

User behaviours after critical mobile application incidents: the relationship with situational context

Users occasionally have critical incidents with information systems (IS). A critical IS incident is an IS product or service experience that a user considers to be unusually positive or negative. Critical IS incidents are highly influential in terms of users' overall perceptions and customer relationships; thus, they are crucial for IS product and service providers. Therefore, it is important to study user behaviours after such incidents. Within IS, the relationships between the situational context and user behaviours after critical incidents have not been addressed at all. Prior studies on general mobile use as a related research area have recognized the influence of the situational context, but they have not covered the relationships between specific situational characteristics and different types of user behaviours. To address this gap, we examine 605 critical mobile incidents that were collected from actual mobile application users. Based on our results, we extend current theoretical knowledge by uncovering and explaining the relationships between specific situational characteristics (interaction state, place, sociality and application type) and user behaviours (use continuance, word‐of‐mouth and complaints). We have found, for example, that users are less likely to engage in negative behaviours after negative incidents that take place outdoors or in vehicles than after indoor incidents. This is because users often consider indoor environments to be familiar and treat them with established expectations and low uncertainty: users are accustomed to the notion that the applications function indoors just like before. Further, we present practical implications for mobile application providers by suggesting to them which positive critical incidents are the most beneficial to promote and which negative critical incidents are the most crucial to avoid.     

Railway-controller-perceived competence in incidents and accidents

Railway controllers play a pivotal role in service recovery of normal rail system operations when incidents and accidents occur. Those in this position must have sufficient competence to overcome task difficulties caused by accident uncertainties. This study adopts Taiwan's railway system as a case study to diagnose railway-controller-perceived competence when facing diverse tasks during incidents and accidents that are derived from a proposed conceptual model. Railway-controller-perceived competence is measured using the Rasch model. Analytical results indicate that working with an external rescue agency handling a rescue operation, explanations to the public, and communication with an external rescue agency are considered the most troublesome tasks. Additionally, railway-controller-perceived competence differs based on the work experience. This information will prove useful for rail system operators and government regulators when designing and regulating railway controller competence management systems. STATEMENT OF RELEVANCE: This study presents a systematic approach for examining the gap between railway-controller-perceived competence and task difficulties associated with incidents and accidents. The relevance of findings encompasses the effects of transportation ergonomics and railway issues on the problem-solving competence and decision-making skills of railway controllers, and the competence management system.     

Railway-controller-perceived competence in incidents and accidents

Railway controllers play a pivotal role in service recovery of normal rail system operations when incidents and accidents occur. Those in this position must have sufficient competence to overcome task difficulties caused by accident uncertainties. This study adopts Taiwan's railway system as a case study to diagnose railway-controller-perceived competence when facing diverse tasks during incidents and accidents that are derived from a proposed conceptual model. Railway-controller-perceived competence is measured using the Rasch model. Analytical results indicate that working with an external rescue agency handling a rescue operation, explanations to the public, and communication with an external rescue agency are considered the most troublesome tasks. Additionally, railway-controller-perceived competence differs based on the work experience. This information will prove useful for rail system operators and government regulators when designing and regulating railway controller competence management systems.

Statement of Relevance: This study presents a systematic approach for examining the gap between railway-controller-perceived competence and task difficulties associated with incidents and accidents. The relevance of findings encompasses the effects of transportation ergonomics and railway issues on the problem-solving competence and decision-making skills of railway controllers, and the competence management system.     

SpinRadar: a spontaneous service provision middleware for place-aware social interactions

With the advancements of mobile phones and the integration of multiple communication interfaces, online social interaction between users is no longer restricted to a specific place with connectivity to the Internet but can happen anywhere and at any time. This has promoted the development of mobile social applications to enable opportunistic interactions with co-located users. One of the challenging problems in such interactions is to discover interaction opportunities with nearby users. Existing works focus on properties related to mobile users in order to find similar users in the surrounding area; these works depend on predefined logic such as conditional statements to recommend spontaneous social interaction opportunities. However, the social implications of the place in which the interaction is taking place are an important factor for recommendations, as those implications provide hints about the most plausible types of interactions among co-located users. In this work, we present a middleware called SpinRadar which is designed to support spontaneous interactions between co-located users by taking into account the semantics of a place, which we call ‘placeness.’ Our evaluation shows that the proposed scheme satisfies users much more than existing schemes.     

Distributed multi-task classification: a decentralized online learning approach

Although dispersing one single task to distributed learning nodes has been intensively studied by the previous research, multi-task learning on distributed networks is still an area that has not been fully exploited, especially under decentralized settings. The challenge lies in the fact that different tasks may have different optimal learning weights while communication through the distributed network forces all tasks to converge to an unique classifier. In this paper, we present a novel algorithm to overcome this challenge and enable learning multiple tasks simultaneously on a decentralized distributed network. Specifically, the learning framework can be separated into two phases: (i) multi-task information is shared within each node on the first phase; (ii) communication between nodes then leads the whole network to converge to a common minimizer. Theoretical analysis indicates that our algorithm achieves a \(\mathcal {O}(\sqrt{T})\) regret bound when compared with the best classifier in hindsight, which is further validated by experiments on both synthetic and real-world datasets.     

Asynchronous broadcast-based decentralized learning in sensor networks

Abstract

In this paper, we study the problem of decentralized learning in sensor networks in which local learners estimate and reach consensus to the quantity of interest inferred globally while communicating only with their immediate neighbours. The main challenge lies in reducing the communication cost in the network, which involves inter-node synchronisation and data exchange. To address this issue, a novel asynchronous broadcast-based decentralized learning algorithm is proposed. Furthermore, we prove that the iterates generated by the developed decentralized method converge to a consensual optimal solution (model). Numerical results demonstrate that it is a promising approach for decentralized learning in sensor networks.     

A robust approach for multi-agent natural resource allocation based on stochastic optimization algorithms

Natural resource allocation is a complex problem that entails difficulties related to the nature of real world problems and to the constraints related to the socio-economical aspects of the problem. In more detail, as the resource becomes scarce relations of trust or communication channels that may exist between the users of a resource become unreliable and should be ignored. In this sense, it is argued that in multi-agent natural resource allocation settings agents are not considered to observe or communicate with each other. The aim of this paper is to study multi-agent learning within this constrained framework. Two novel learning methods are introduced that operate in conjunction with any decentralized multi-agent learning algorithm to provide efficient resource allocations. The proposed methods were applied on a multi-agent simulation model that replicates a natural resource allocation procedure, and extensive experiments were conducted using popular decentralized multi-agent learning algorithms. Experimental results employed statistical figures of merit for assessing the performance of the algorithms with respect to the preservation of the resource and to the utilities of the users. It was revealed that the proposed learning methods improved the performance of all policies under study and provided allocation schemes that both preserved the resource and ensured the survival of the agents, simultaneously. It is thus demonstrated that the proposed learning methods are a substantial improvement, when compared to the direct application of typical learning algorithms to natural resource sharing, and are a viable means of achieving efficient resource allocations.     

Recommending items in pervasive scenarios: models and experimental analysis

In this paper, we propose and investigate the effectiveness of fully decentralized, collaborative filtering techniques. These are particularly interesting for use in pervasive systems of small devices with limited communication and computational capabilities. In particular, we assume that items are tagged with smart tags (such as passive RFIDs), storing aggregate information about the visiting patterns of users that interacted with them in the past. Users access and modify information stored in smart tags transparently, by smart reader devices that are already available on commercial mobile phones. Smart readers use private information about previous behavior of the user and aggregate information retrieved from smart tags to recommend new items that are more likely to meet user expectations. Note that we do not assume any transmission capabilities between smart tags: Information exchange among them is mediated by users’ collective and unpredictable navigation patterns. Our algorithms do not require any explicit interaction among users and can be easily and efficiently implemented. We analyze their theoretical behavior and assess their performance in practice, by simulation on both synthetic and real, publicly available data sets. We also compare the performance of our fully decentralized solutions with that of state-of-the-art centralized strategies.     

What Is Learning? A Review of the Safety Literature to Define Learning from Incidents,Accidents and Disasters

Learning from incidents, accidents and disasters contributes to improvement of safety and the prevention of unwanted events. In this review, literature on learning from safety incidents within organizations is studied and compared with the organizational learning theory of Argyris and Schön. Sub‐processes, such as learning lessons, sharing, storing and applying lessons, are described, and factors that influence these processes are listed, such as trust, the severity of the consequences of an incident and the people involved in learning. In comparison with the theory of Argyris and Schön, aspects about the information to learn from, i.e., the incident and analysis, are much more specified in the safety literature. However, the organizational learning theory gives more details about the earlier mentioned sub‐processes.     

Internet of intelligent things and robot as a service

The development of computing and communication systems has gone through a spiral cycle of centralization and decentralization paradigms. The earliest computer systems are centralized mainframe computers. The paradigm moved to decentralized as networked stations became more dependable, extensible and cost-effective. The decentralized systems have their limitations and inconveniences. The virtualization and cloud computing paradigm creates a centralized system that appears to users to be a centralized system, where computing and communication resources are not in the client computers but in an integrated infrastructure that is accessible anywhere and anytime. Nevertheless, the implementation of the centralized infrastructure is equipped with decentralized and redundant resources, which makes the system more dependable as any component failures can be tolerated internally. The Internet of Things extends the cloud computing concept beyond computing and communication to include everything, particularly, the physical devices. This paper discusses the architectures, interfaces, and behaviors of intelligent devices connected to the cloud computing environment. Robot as a Service is the case study, which has all the key features of Internet of Intelligent Things: autonomous, mobile, sensing, and action taking. The goal is to further extend the centralized cloud computing environment into a decentralized system to complete another cycle of the spiral development. The idea of achieving the goal is through autonomous and intelligent mobile physical services or robots as services to form local pool of intelligent devices and that can make local decisions without communicate with the cloud.     

基于频谱合约的协作频谱共享方法

针对授权频谱资源不足的问题,结合非对称网络信息的特点和通信双方自私性的需求,提出一种多用户协作频谱共享(CSS)合约机制的设计与实现方法。首先,通过将CSS映射成劳动力市场,研究对称网络信息条件下多用户频谱合约建模方法;在此基础上,针对合约签订后非授权用户(SU)行为隐匿性引起的道德风险问题,研究非对称网络信息条件下频谱合约建模方法,以激励双方协作行为和保证频谱共享的实现。实验结果表明,在授权用户(PU)直接传输速率低于0.2 b/s时,协作频谱共享时整个系统容量是双方不参与共享时的3倍以上。提出的多用户CSS合约机制能为无线频谱的高效利用和资源共享提供新思路。     

The user–developer communication process: a critical case study

Abstract. Although user participation in systems development is widely believed to have positive impacts on user acceptance, it does not guarantee success and there is still much that we do not know about how and why user participation sometimes delivers positive benefits, but not always. Much of the prior research on user participation assumes that user–developer communication will ensure that the resulting system will be designed to meet users' needs and will be accepted by them. The nature and quality of the communication between users and developers, however, remains an understudied aspect of user participation. In this paper, we focus on the user–developer communication process. We propose a process model that delineates four stages of communication between users and software developers, and we argue that these stages must occur for user participation to lead to effective outcomes. To illustrate our model, we apply it to analyse a 'critical case study' of a software project that failed despite high levels of user involvement. We show that when 'communication lapses' occurred in several of the user–developer communication stages, developers failed to be informed regarding the underlying reasons that users avoided the system. Based on the insights from this case study, we advise researchers and practitioners how to leverage the potential benefits of user participation, rather than take them for granted.     

Risky behaviour at traffic signals: a traffic engineer's view

Abstract

Road accidents are reasonably predictable from a knowledge of traffic behaviour, although individual accident involvement is far less predictable. How do individuals perceive the relationship between their own involvement and objective accident risk? This question is explored in relation to Danish case studies of drivers' behaviour at traffic lights and different categories of pedestrians' ‘jay-walking’. On this evidence, it seems unlikely that road users perceive accidents as random negative outcomes of everyday risk taking. Thus it also appears improbable that overall traffic accident risk in any society is a major function of deliberate risk taking by its individual road users.     

Dynamic pricing and channel efficiency in the presence of the cost learning effect

This paper investigates the impacts of the supply‐side cost learning effect on dynamic pricing strategies and the channel efficiency in a decentralized supply chain consisting of a manufacturer and a retailer. The unit production cost incurred by the manufacturer declines as the cumulative output increases. With the manufacturer being the price leader, the interactions between the two channel members are modeled as a Stackelberg differential game, where the two channel members dynamically set the wholesale and retail prices over time to optimize their respective profits. We find that while the cost learning effect is beneficial to both channel members in terms of long‐term profitability, such an effect is detrimental to the channel inefficiency caused by price double marginalization. In a setting where the manufacturer sells through competing retailers, we show that the channel efficiency can be improved with the emergence of horizontal competition. Our result also indicates that the channel efficiency with a feedback pricing equilibrium is higher than that with an open‐loop equilibrium.     

Analytical model of adoption of item level RFID in a two-echelon supply chain with shelf-space and price-dependent demand

We investigate whether it is possible for the manufacturer as well as the retailer to derive economic benefits from item-level RFID. We consider two supply chain contracts. In the first instance the manufacturer sets the wholesale price for his product and in response the retailer sets the retail price and allocates the shelf-space. We find that in this setting item-level RFID is adopted in a decentralized supply chain at the same levels of RFID costs as in an integrated supply chain with a single decision maker. The second case that we consider involves a powerful retailer setting product shelf-space prices. In this setting the manufacturer essentially rents shelf-space from the retailer. We find that in this setting, the retailer may force the adoption of RFID even if it does not maximize the chain's total profit.     

ARREST: From work practices to redesign for usability

In this paper we discuss the redesign of a support management system deployed in a small and medium sized enterprise (SME) in the UK. The original system was not fulfilling its needs as it had not captured work practices in a way that was recognizable to the users. The advantages of the redesign included: improved usefulness; improved efficiency and productivity; reduced learning time; improved usability; and increased acceptance among users. The system is used to support complex and distributed cooperative activities taking place in an SME. We evaluated the current system and analysed work practices using a user-centred design and evaluation philosophy. In this paper we discuss how user needs are incorporated into the enhanced design of the support management system. The user-centred design techniques used in this research include interviews, questionnaires, observations and user tests. We present comparative evaluation results that show significant improvement in performance of user tasks using the redesigned support management system. The contribution of this paper is the presentation of a case study to show how a user-centred design and evaluation philosophy can lead to better requirements capture resulting in systems that more accurately capture the users’ conceptual models.     

Learning from Workaround Practices: the Challenge of Enterprise System Implementations in Multinational Corporations

Based on a four‐year longitudinal case study of a French multinational corporation (MNC) this research explains how a global enterprise system, initially designed to provide greater control by headquarters across its MNC sites, led to the emergence of unexpected practices – known as workarounds – in its Chinese subsidiaries. Drawing on concepts from activity theory our findings provide greater theoretical understanding of workarounds in three ways. First, users are involved in collectively constructing and implementing workaround practices rather than simply being passive users or reduced to the dichotomous framing of ‘accepting’ or ‘rejecting’ an information system. Workarounds are not simple individual user improvizations but are collectively developed and become unofficial local rules. Second, workarounds are part of learning which goes beyond the view of workarounds as deviant and non‐compliant behavior. Third, learning activities are not restricted to users and their practices. Rather, as we demonstrate learning and evolution take place between the users, managers, processes and the system; that is, workarounds are not only a critical part of ‘getting work done’ but also an integral part of the institutionalization of an enterprise system.     

The impact of IT on decision structure and firm performance: evidence from the textile and apparel industry

Clarifying the relationships between information technology (IT), organizational performance, and decision structure remains an important area of inquiry in IS research. Through an empirical analysis and complementary case examples, our study examines these associations among firms operating in the US-based apparel and textile industry from 1992 to 1997. Based on data gathered from 50 public firms located across the USA, the study finds that IT used to enhance internal communication supports a decentralized decision structure, which in turn is associated with higher financial performance. Hence, IT exhibits an indirect performance effect. However, use of IT to enhance communication is also found to have a direct performance effect in large organizations. This paper proposes that use of communication enhancing IT can support organizational learning processes by facilitating flexible exchange of skills and knowledge across functional areas. Case examples are used to illustrate how these learning effects can materialize.     

Generic service integration in adaptive learning experiences using IMS learning design

IMS Learning Design is a specification to capture the orchestration taking place in a learning scenario. This paper presents an extension called Generic Service Integration. This paradigm allows a bidirectional communication between the course engine in charge of the orchestration and conventional Web 2.0 tools. This communication allows the engine to configure external services so as to adjust their behaviour, and also retrieve information about their activity and use it to influence the orchestration process. This extension impacts all the phases of the life cycle of a learning activity: authoring, deployment and enactment. The paper analyses this impact and proposes appropriate adjustments. A pilot test was conducted and the obtained results show that learning experiences that adapt their flow-based on third party tools while maintaining interoperability, reusability and self-containment requirements are feasible.