Image quality assessment (IQA) and its computational models play a vital role in modern computer vision applications. Research has traditionally focused on signal distortions arising during image compression and transmission, and their impact on perceived image quality. However, little attention is paid to image manipulation that alters an image using various filters. With the prevalence of image manipulation in real-life scenarios, it is critical to understand how humans perceive filter-altered images and to develop reliable IQA models capable of automatically assessing the quality of filtered images. In this paper, we build a new IQA database for filter-altered images, comprised of 360 images manipulated by various filters. To ensure the subjective IQA faithfully reflects human visual perception, we conduct a fully-controlled psychovisual experiment. Building upon the ground truth, we propose an innovative deep learning-based no-reference IQA (NR-IQA) model named IMQA that can accurately predict the perceived quality of filter-altered images. This model involves constructing an image filtering-aware module to learn discriminatory features for filter-altered images; and fuses these features with the representations generated by an image quality-aware module. Experimental results demonstrate the superior performance of the proposed IMQA model.

The push to a connected world where people carry an always-online device which has been designed to maximise instant gratification and prompts users via notifications has lead to a surge of potentially problematic behaviour as a result. This has lead to a rising interest in addressing and understanding the addictiveness of smartphone usage, as well as for particular applications (apps). However, capturing addiction from usage involves not only assessment of potential addiction risk but also requires understanding of the complex interactions that define user behaviour and how these can be effectively isolated and summarised. In this paper, we examine the correlation of physical user interface (UI) interactions (e.g. taps and scrolls) and smartphone addiction risk using a large dataset of those smartphone events (65,093,343, N=301,024 sessions) collected from 64 users over an 8-week period with an accompanying smartphone addiction survey. Our novel method which reports on the probability of a users addiction risk and in a model case we show how it was be used to identify 57 of 64 users correctly. This supports our observations of UI events during sessions of usage being indicative of addiction risk while improving previous approaches which rely on summative data such as screen on time. Within this we also find that users only exhibit addictive behaviour in a subset of all sessions while using their smartphone.

The prevalence of induced triads play an important role in characterising complex networks, supporting approaches for assessment of dynamic and partially obfuscated scenarios. In this paper we introduce a new local edge-centrality measure that is designed to be deployed in this context for complex networks and is highly scalable. It signifies the importance an edge plays within induced triads for a directed network. We observe that an edge can play one of two roles in providing connectivity within any particular triad, based on whether the edge supports connectivity to the third node or not. We call these alternative states overt and covert. As an edge may play alternative roles in different induced triads, this allows us to assess the local importance of an edge across multiple induced substructures. We introduce theory to count the number of induced triads in which an edge is overt and covert. Using 34 data sets derived from public sources, we show how the presence of overt and covert edges can be used to profile diverse real-world networks. The relationship with global network analysis metrics is examined. We observe that overt and covert edge centrality is useful in further differentiating classes of network, when considered in combination with conventional global network analysis metrics.

SUMMARY: We present LipidFinder 2.0, incorporating four new modules that apply artefact filters, remove lipid and contaminant stacks, in-source fragments and salt clusters, and a new isotope deletion method which is significantly more sensitive than available open-access alternatives. We also incorporate a novel false discovery rate method, utilizing a target-decoy strategy, which allows users to assess data quality. A renewed lipid profiling method is introduced which searches three different databases from LIPID MAPS and returns bulk lipid structures only, and a lipid category scatter plot with color blind friendly pallet. An API interface with XCMS Online is made available on LipidFinder's online version. We show using real data that LipidFinder 2.0 provides a significant improvement over non-lipid metabolite filtering and lipid profiling, compared to available tools. AVAILABILITY AND IMPLEMENTATION: LipidFinder 2.0 is freely available at https://github.com/ODonnell-Lipidomics/LipidFinder and http://lipidmaps.org/resources/tools/lipidfinder. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

SUMMARY: We present LipidFinder online, hosted on the LIPID MAPS website, as a liquid chromatography/mass spectrometry (LC/MS) workflow comprising peak filtering, MS searching and statistical analysis components, highly customized for interrogating lipidomic data. The online interface of LipidFinder includes several innovations such as comprehensive parameter tuning, a MS search engine employing in-house customized, curated and computationally generated databases and multiple reporting/display options. A set of integrated statistical analysis tools which enable users to identify those features which are significantly-altered under the selected experimental conditions, thereby greatly reducing the complexity of the peaklist prior to MS searching is included. LipidFinder is presented as a highly flexible, extensible user-friendly online workflow which leverages the lipidomics knowledge base and resources of the LIPID MAPS website, long recognized as a leading global lipidomics portal. AVAILABILITY AND IMPLEMENTATION: LipidFinder on LIPID MAPS is available at: http://www.lipidmaps.org/data/LF.

Detecting communities that recur over time is a challenging problem due to the potential sparsity of encounter events at an individual scale and inherent uncertainty in human behavior. Existing methods for community detection in mobile human encounter networks ignore the presence of temporal patterns that lead to periodic components in the network. Daily and weekly routine are prevalent in human behavior and can serve as rich context for applications that rely on person-to-person encounters, such as mobile routing protocols and intelligent digital personal assistants. In this article, we present the design, implementation, and evaluation of an approach to decentralized periodic community detection that is robust to uncertainty and computationally efficient. This alternative approach has a novel periodicity detection method inspired by a neural synchrony measure used in the field of neurophysiology. We evaluate our approach and investigate human periodic encounter patterns using empirical datasets of inferred and direct-sensed encounters.

Device-to-device (D2D) communications are recognized as a key enabler of future cellular networks, which will help to drive improvements in spectral efficiency and assist with the offload of network traffic. Relay-assisted D2D communications will be essential when there is an extended distance between the source and the destination or when the transmit power is constrained below a certain level. Although a number of works on relay-assisted D2D communications have been presented in the literature, most of those assume that relay nodes cooperate unequivocally. In reality, this cannot be assumed, since there is little incentive to cooperate without a guarantee of future reciprocal behavior. To incorporate the social behavior of D2D nodes, we consider the decision to relay using the donation game based on social comparison, characterize the probability of cooperation in an evolutionary context and then evaluate the network performance of relay-assisted D2D communications. Through numerical evaluations, we investigate the performance gap between the ideal case of 100% cooperation and practical scenarios with a lower cooperation probability. It shows that practical scenarios achieve lower transmission capacity and higher outage probability than idealistic network views, which assume full cooperation. After a sufficient number of generations, however, the cooperation probability follows the natural rules of evolution and the transmission performance of practical scenarios approach that of the full cooperation case, indicating that all D2D relay nodes adapt the same dominant cooperative strategy based on social comparison, without the need for external enforcement.

Social networks are known to form on the basis of homophily, where nodes with some type of similar characteristics are more likely to be connected. Some of the most fundamental human characteristics are reflected by an individual's personality, which represents a persistent disposition governing a human's outlook and approach to diverse situations. While taking into account demographics of age and gender, we assess the extent to which personality homophily is evident in the local network features of Facebook. Using a large sample obtained from the MyPersonality dataset, we find that a range of network-based features correlate with personality facets of individuals. In particular, extraversion had a positive effect on an individual's network size, while neuroticism had a negative effect. Additionally, extraversion and openness were positively related to transitivity, which was moderated by gender. Finally, we found that conscientiousness, agreeableness and extraversion were homophilous: people with higher similarity on these facets were more strongly connected. This was additionally mediated by gender for agreeableness: personality similarity had an effect for male-only and mixed pairs, but not for female-only pairs. Personality similarity was also stronger among closed triangles, compared to open ones. These results support the idea that inherent attraction between individuals, on the basis of personality, drives the roles we play within our online social networks.

Device-to-device (D2D) communications are recognized as a key component of future wireless networks which will help to improve spectral efficiency and network densification simultaneously. In order to guarantee a quality of service (QoS) to the cellular links, the transmit power of the D2D nodes needs to be restricted, which has lead to a poor link quality over D2D transmission. One viable option to improve the D2D link quality is incorporating cooperative relays into D2D networks. However most of the existing published work in relay assisted D2D networks has assumed that relay nodes cooperate spontaneously. This cannot always be guaranteed and we take this into account by considering a fundamental model on which donation-based cooperation depends. In particular we model relay cooperation as a donation game based on social comparison and characterize cooperation probability in an evolutionary context. When applying this model we evaluate the outage and capacity of relay assisted D2D network using a stochastic geometric framework.

Online social networks such as Twitter have emerged as an important mechanism for individuals to share information and post user generated content . However, filtering interesting content from the large volume of messages received through Twitter places a significant cognitive burden on users. Motivated by this problem, we develop a new automated mechanism to detect personalised interestingness, and investigate this for Twitter. Instead of undertaking semantic content analysis and matching of tweets, our approach considers the human response to content, in terms of whether the content is sufficiently stimulating to get repeatedly chosen by users for forwarding (retweeting). This approach involves machine learning against features that are relevant to a particular user and their network, to obtain an expected level of retweeting for a user and a tweet. Tweets observed to be above this expected level are classified as interesting. We implement the approach in Twitter and evaluate it using comparative human tweet assessment in two forms: through aggregated assessment using Mechanical Turk , and through a web-based experiment for Twitter users. The results provide confidence that the approach is effective in identifying the more interesting tweets from a user’s timeline. This has important implications for reduction of cognitive burden: the results show that timelines can be considerably shortened while maintaining a high degree of confidence that more interesting tweets will be retained. In conclusion we discuss how the technique could be applied to mitigate possible filter bubble effects.

When should a machine attempt to communicate with a user? This is a historical problem that has been studied since the rise of personal computing. More recently, the emergence of pervasive technologies such as the smartphone have extended the problem to be ever-present in our daily lives, opening up new opportunities for context awareness through data collection and reasoning. Complementary to this there has been increasing interest in techniques to intelligently synchronise interruptions with human behaviour and cognition. However, it is increasingly challenging to categorise new developments, which are often scenario specific or scope a problem with particular unique features. In this paper we present a meta-analysis of this area, decomposing and comparing historical and recent works that seek to understand and predict how users will perceive and respond to interruptions. In doing so we identify research gaps, questions and opportunities that characterise this important emerging field for pervasive technology.

Social micro-blogging systems such as Twitter are designed for rapid and informal communication from a large potential number of participants. Due to the volume of content received, human users must typically skim their timeline of received content and exercise judgement in selecting items for consumption, necessitating a selection process based on heuristics and content meta-data. This selection process is not well understood, yet is important due to its potential use in content management systems. In this research we have conducted an open online experiment in which participants are shown quantitative and qualitative meta-data describing two pieces of Twitter content. Without revealing the text of the tweet, participants are asked to make a selection. We observe the decisions made from 239 surveys and discover insights into human behaviour on decision making for content selection. We find that for qualitative meta-data consumption decisions are driven by online friendship and for quantitative meta-data the largest numerical value presented influences choice. Overall, the ‘number of retweets’ is found to be the most influential quantitative meta-data, while displaying multiple cues about an author׳s identity provides the strongest qualitative meta-data. When both quantitative and qualitative meta-data is presented, it is the qualitative meta-data (friendship information) that drives selection. The results are consistent with application of the Recognition heuristic, which postulates that when faced with constrained decision-making, humans will tend to exercise judgement based on cues representing familiarity. These findings are useful for future interface design for content filtering and recommendation systems.

Until recently crowdsourcing has been primarily conceived as an online activity to harness resources for problem solving. However, the emergence of Opportunistic Networking (ON) has opened up crowdsourcing to the spatial domain. In this article, we bring the ON model for potential crowdsourcing in the smart city environment. We introduce cognitive features of the ON that allow users’ mobile devices to become aware of the surrounding physical environment. Specifically, we exploit cognitive psychology studies on dynamic memory structures and cognitive heuristics—mental models that describe how the human brain handles decision making among complex and real-time stimuli. Combined with ON, these cognitive features allow devices to act as proxies in their users’ cyberworlds and exchange knowledge to deliver awareness of places in an urban environment. This is done through tags associated with locations. They represent features that are perceived by humans about a place. We consider the extent to which this knowledge becomes available to participants using interactions with locations and other nodes. This is assessed taking into account a wide range of cognitive parameters. Outcomes are important because this functionality could support a new type of recommendation system that is independent of the traditional forms of networking.

Many wireless devices employ multi-rate techniques to improve network performance. However, despite the significant amount of research aimed at dynamically adjusting the transmission rate, the majority of this effort considers neither the competing nodes in wireless mesh networks nor the congestion in the nodes. This work employs distributed intelligent agents to observe the surrounding environment in order to dynamically adjust the individual node transmission rates. Reinforcement learning is employed to control the way each node updates its transmission rate based on the transmission rate of the adjacent node as well as the traffic load. This work is validated through extensive simulations that compare the proposed model with three of the most widely cited schemes. The results indicate significant improvement in system throughput.

Micro-blogging services have recently been experiencing increasing success among Web users. Different to traditional online social applications, micro-blogs are lightweight, require small cognitive effort and help share real-time information about personal activities and interests. In this article, we explore scalable pushing protocols that are particularly suited for the delivery of this type of service in a mobile pervasive environment. Here, micro-blog updates are generated and carried by mobile (smart-phone type) devices and are exchanged through opportunistic encounters. We enhance primitive push mechanisms using social information concerning the interests of network nodes as well as the frequency of encounters with them. This information is collected and shared dynamically, as nodes initially encounter each other and exchange their preferences, and directs the forwarding of micro-blog updates across the network. Also incorporated is the spatiotemporal scope of the updates, which is only partially considered in current Internet services. We introduce several new protocol variants that differentiate the forwarding strategy towards interest-similar and frequently encountered nodes, as well as the amount of updates forwarded upon each encounter. In all cases, the proposed scheme outperforms the basic flooding dissemination mechanism in delivering high numbers of micro-blog updates to the nodes interested in them. Our extensive evaluation highlights how use can be made of different amounts of social information to trade performance with complexity and computational effort. However, hard performance bounds appear to be set by the level of coincidence between interest-similar node communities and meeting groups emerging due to the mobility patterns of the nodes.

Epidemic type exchange protocols have often been discussed as a means of disseminating information through a network of nodes. These protocols may be applied to mobile opportunistic ad-hoc networks where the goal is not total dissemination of all information to all nodes but where nodes are interested in attaining specific content for which they have a particular need or desire. In this proposed scenario a shuffle exchange protocol is inefficient, requiring many unnecessary exchanges in order for nodes to receive the content they require. Allowing nodes to self-adapt the amount of content exchanged with each neighbour based on the quality of information received makes it possible for nodes to efficiently receive all content in which they are interested.

In this paper, we consider the optimisation of transmission schedules for infrastructure Wireless Mesh Networks in which data is forwarded through mesh routers from a single Internet Gateway node. The mesh routers receive and aggregate data from local mobile devices and each mesh router has an assigned data allowance to ensure fairness, set depending on its geographical position or the predicted usage patterns. We examine the use of fair and efficient link scheduling for Wireless Mesh Networks and provide an integer program for maximising the throughput allowance for each mesh router in a network given the topology. The program uses a slotted time approach to maximise the throughput within a given number of slots N, thus allowing a network to be split into sub networks for local access to the mesh routers, and back-haul transmissions to the gateway. Results are presented showing the optimised throughput for a selection of networks and a range of values for N.

We address the problem of cooperation in decentralized systems, specifically looking at interactions between independent pairs of peers where mutual exchange of resources (e.g., updating or sharing content) is required. In the absence of any enforcement mechanism or protocol, there is no incentive for one party to directly reciprocate during a transaction with another. Consequently, for such decentralized systems to function, protocols for self-organization need to explicitly promote cooperation in a manner where adherence to the protocol is incentivized. In this article we introduce a new generic model to achieve this. The model is based on peers repeatedly interacting to build up and maintain a dynamic social network of others that they can trust based on similarity of cooperation. This mechanism effectively incentivizes unselfish behavior, where peers with higher levels of cooperation gain higher payoff. We examine the model's behavior and robustness in detail. This includes the effect of peers self-adapting their cooperation level in response to maximizing their payoff, representing a Nash-equilibrium of the system. The study shows that the formation of a social network based on reflexive cooperation levels can be a highly effective and robust incentive mechanism for autonomous decentralized systems.

This paper applies a problem decomposition approach in order to solve hard Frequency Assignment Problem instances with standard meta-heuristics. The proposed technique aims to divide the initial problem into a number of easier subproblems, which can then be solved either independently or in sequence respecting the constraints between them. Finally, partial subproblems solutions are recomposed into a solution of the original problem. Our results focus on the COST-259 MI-FAP instances, for which some good assignments produced by local search meta-heuristics are widely available. However, standard implementations do not usually produce the best performance and, in particular, no good results have been previously obtained using evolutionary techniques. We show that problem decomposition can improve standard heuristics, both in terms of solution quality and runtime. Furthermore, genetic algorithms seem to benefit more from this approach, showing a higher percentage improvement, therefore reducing the gap with other local search methods.

We investigate the automatic design of efficient wireless communications networks whose nodes exhibit mobility characteristics and where a subset of the nodes must be selected to host servers. The paper describes a representation of the problem, an evolutionary solution framework and some preliminary results.