A multi-view framework for generating mobile apps

This paper demonstrates a multi-view framework for Rapid APPlication Tool (RAPPT). RAPPT enables rapid development of mobile applications. It employs a multilevel approach to mobile application development: a Domain Specific Visual Language to define the high level structure of mobile apps, a Domain Specific Textual Language to define behavioural concepts, and concrete source code for fine grained improvements.

Predictors of pre- and post-competition affective states in male martial artists : a multilevel interactional approach

The aims of this study were to examine (a) the effects of competition-related and competition-extraneous concerns on affective states; (b) the relationships of primary and secondary appraisal with affective states and (c) the main and moderating effects of personality traits on pre- and post-competition affects. Thirty-nine male elite martial artists were assessed on 12 affective states, concerns and dimensions of primary and secondary appraisal at five random times a day across 1 week before and 3 days after a competition. On the competition day, they were assessed 1 h before and immediately after the contest. Competitive trait anxiety, neuroticism and extraversion were measured at the start of the study. The competition was the most significant and stressful event experienced in the examined period and had a pervasive influence on athletes' affective states. All examined appraisal and personality factors were somewhat associated with pre- and post-competition affective states. Competitive trait anxiety was a key moderator of the relationship between cognitive appraisal and affective states. This study supports the idea that cognitive appraisal and situational and personality factors exert main and interactive effects on athletes' pre- and post-competition affects. These factors need to be accounted for in planning of emotion regulation interventions.

Modelling multilevel data in multimedia : A hierarchical factor analysis approach

Multimedia content understanding research requires rigorous approach to deal with the complexity of the data. At the crux of this problem is the method to deal with multilevel data whose structure exists at multiple scales and across data sources. A common example is modeling tags jointly with images to improve retrieval, classification and tag recommendation. Associated contextual observation, such as metadata, is rich that can be exploited for content analysis. A major challenge is the need for a principal approach to systematically incorporate associated media with the primary data source of interest. Taking a factor modeling approach, we propose a framework that can discover low-dimensional structures for a primary data source together with other associated information. We cast this task as a subspace learning problem under the framework of Bayesian nonparametrics and thus the subspace dimensionality and the number of clusters are automatically learnt from data instead of setting these parameters a priori. Using Beta processes as the building block, we construct random measures in a hierarchical structure to generate multiple data sources and capture their shared statistical at the same time. The model parameters are inferred efficiently using a novel combination of Gibbs and slice sampling. We demonstrate the applicability of the proposed model in three applications: image retrieval, automatic tag recommendation and image classification. Experiments using two real-world datasets show that our approach outperforms various state-of-the-art related methods.

A Bayesian nonparametric approach to multilevel regression

Regression is at the cornerstone of statistical analysis. Multilevel regression, on the other hand, receives little research attention, though it is prevalent in economics, biostatistics and healthcare to name a few. We present a Bayesian nonparametric framework for multilevel regression where individuals including observations and outcomes are organized into groups. Furthermore, our approach exploits additional group-specific context observations, we use Dirichlet Process with product-space base measure in a nested structure to model group-level context distribution and the regression distribution to accommodate the multilevel structure of the data. The proposed model simultaneously partitions groups into cluster and perform regression. We provide collapsed Gibbs sampler for posterior inference. We perform extensive experiments on econometric panel data and healthcare longitudinal data to demonstrate the effectiveness of the proposed model

Multilevel analysis of the Be Active Eat Well intervention : environmental and behavioural influences on reductions in child obesity risk

Background: The Be Active Eat Well (BAEW) community-based child obesity prevention intervention was successful in modestly reducing unhealthy weight gain in primary school children using a multi-strategy and multi-setting approach.

Objective: To (1) examine the relationship between changes in obesity-related individual, household and school factors and changes in standardised child body mass index (zBMI), and (2) determine if the BAEW intervention moderated these effects.

Methods: The longitudinal relationships between changes in individual, household and school variables and changes in zBMI were explored using multilevel modelling, with measurement time (baseline and follow-up) at level 1, individual (behaviours, n=1812) at level 2 and households (n=1318) and schools (n=18) as higher levels (environments). The effect of the intervention was tested while controlling for child age, gender and maternal education level.

Results: This study confirmed that the BAEW intervention lowered child zBMI compared with the comparison group (−0.085 units, P=0.03). The variation between household environments was found to be a large contributor to the percentage of unexplained change in child zBMI (59%), compared with contributions from the individual (23%) and school levels (1%). Across both groups, screen time (P=0.03), sweet drink consumption (P=0.03) and lack of household rules for television (TV) viewing (P=0.05) were associated with increased zBMI, whereas there was a non-significant association with the frequency the TV was on during evening meals (P=0.07). The moderating effect of the intervention was only evident for the relationship between the frequency of TV on during meals and zBMI, however, this effect was modest (P=0.04).

Conclusions: The development of childhood obesity involves multi-factorial and multi-level influences, some of which are amenable to change. Obesity prevention strategies should not only target individual behaviours but also the household environment and family practices. Although zBMI changes were modest, these findings are encouraging as small reductions can have population level impacts on childhood obesity levels.

A multilevel social network perspective on IT adoption

Adoption of technologies has long been a key area of research in the information systems (IS) discipline, and researchers have thus been interested in the attributes, beliefs, intentions, and behaviors of individuals and organisations that could explain information and communication technology (ICT) adoption. The focal unit of adoption has mainly been individuals and organisations, however, research at group or social network level as well as the inter-organizational level have recently gained increased interest from IS researchers. This recent focus supports the view of the world as being the sum of all relations. Various social network theories exist that seek to emphasize different proficiencies of social networks and explain theoretical mechanisms for behavior in social networks. The core idea of these theories is that social networks are valuable, and the relations among actors affect the behavior of individuals, groups, organizations, industries, and societies. IS researchers have also found that social network theory can help explain technology adoption. Some researchers, in addition, acknowledge that most adoption situations involve phenomena occuring at multiple levels, yet most technology adoption research applies a single level of analysis. Multilevel research can address the levels of theory, measurement, and analysis required to fully examining research questions. This paper therfore adapts the Coleman diagram into the Multi-level Framework of Technology. Adoption in order to explain how social network theory, at the individual and social network level, can help explain adoption of ICT. As Coleman (1990) attempts to create a link between the micro and macro level in a holistic manner, his approach is applicable in explaining ICT adoption

A multilevel trust management framework for service oriented environment

In service-oriented computing applications, trust management systems are emerging as a promising technology to improve the e-commerce consumers and provider's relationship. Both consumers and providers need to evaluate the trust levels of potential partners before engaging in interactions. The accuracy of trust evaluation greatly affects the success rate of the interaction. This paper addresses the threats and challenges that can compromise the reliability of the current trust management system. This paper studies and examines the importance of the trust factors of the trust management framework, specifically in dealing with malicious feedback ratings from e-commerce users. To improve the reliability of the trust management systems, an approach that addresses feedback-related vulnerabilities is paramount. A multilevel trust management system computes trust by combining different types of information. Using this combination, we introduce a multilevel framework for a new interactive trust management to improve the correctness in estimate of trust information.

Scalable nonparametric Bayesian multilevel clustering

Multilevel clustering problems where the con-tent and contextual information are jointly clustered are ubiquitous in modern datasets. Existing works on this problem are limited to small datasets due to the use of the Gibbs sampler. We address the problem of scaling up multi-level clustering under a Bayesian nonparametric setting, extending the MC2 model proposed in (Nguyen et al., 2014). We ground our approach in structured mean-field and stochastic variational inference (SVI) and develop a tree-structured SVI algorithm that exploits the interplay between content and context modeling. Our new algorithm avoids the need to repeatedly go through the corpus as in Gibbs sampler. More crucially, our method is immediately amendable to parallelization, facilitating a scalable distributed implementation on the Apache Spark platform. We conduct extensive experiments in a variety of domains including text, images, and real-world user application activities. Direct comparison with the Gibbs-sampler demonstrates that our method is an order-of-magnitude faster without loss of model quality. Our Spark-based implementation gains an-other order-of-magnitude speedup and can scale to large real-world datasets containing millions of documents and groups.