A two-stage item recommendation method using probabilistic ranking with reconstructed tensor model

In a tag-based recommender system, the multi-dimensional item, tag> correlation should be modeled effectively for finding quality recommendations. Recently, few researchers have used tensor models in recommendation to represent and analyze latent relationships inherent in multi-dimensions data. A common approach is to build the tensor model, decompose it and, then, directly use the reconstructed tensor to generate the recommendation based on the maximum values of tensor elements. In order to improve the accuracy and scalability, we propose an implementation of the -mode block-striped (matrix) product for scalable tensor reconstruction and probabilistically ranking the candidate items generated from the reconstructed tensor. With testing on real-world datasets, we demonstrate that the proposed method outperforms the benchmarking methods in terms of recommendation accuracy and scalability.

A tag-based personalized item recommendation system using tensor modeling and topic model approaches

This research falls in the area of enhancing the quality of tag-based item recommendation systems. It aims to achieve this by employing a multi-dimensional user profile approach and by analyzing the semantic aspects of tags. Tag-based recommender systems have two characteristics that need to be carefully studied in order to build a reliable system. Firstly, the multi-dimensional correlation, called as tag assignment item, tag>, should be appropriately modelled in order to create the user profiles [1]. Secondly, the semantics behind the tags should be considered properly as the flexibility with their design can cause semantic problems such as synonymy and polysemy [2]. This research proposes to address these two challenges for building a tag-based item recommendation system by employing tensor modeling as the multi-dimensional user profile approach, and the topic model as the semantic analysis approach. The first objective is to optimize the tensor model reconstruction and to improve the model performance in generating quality rec-ommendation. A novel Tensor-based Recommendation using Probabilistic Ranking (TRPR) method [3] has been developed. Results show this method to be scalable for large datasets and outperforming the benchmarking methods in terms of accuracy. The memory efficient loop implements the n-mode block-striped (matrix) product for tensor reconstruction as an approximation of the initial tensor. The probabilistic ranking calculates the probabil-ity of users to select candidate items using their tag preference list based on the entries generated from the reconstructed tensor. The second objective is to analyse the tag semantics and utilize the outcome in building the tensor model. This research proposes to investigate the problem using topic model approach to keep the tags nature as the “social vocabulary” [4]. For the tag assignment data, topics can be generated from the occurrences of tags given for an item. However there is only limited amount of tags availa-ble to represent items as collection of topics, since an item might have only been tagged by using several tags. Consequently, the generated topics might not able to represent the items appropriately. Furthermore, given that each tag can belong to any topics with various probability scores, the occurrence of tags cannot simply be mapped by the topics to build the tensor model. A standard weighting technique will not appropriately calculate the value of tagging activity since it will define the context of an item using a tag instead of a topic.

Online estimation of Allan variance parameters

A new online method is presented for estimation of the angular randomwalk and rate randomwalk coefficients of inertial measurement unit gyros and accelerometers. In the online method, a state-space model is proposed, and recursive parameter estimators are proposed for quantities previously measured from offline data techniques such as the Allan variance method. The Allan variance method has large offline computational effort and data storage requirements. The technique proposed here requires no data storage and computational effort of approximately 100 calculations per data sample.

Reduced complexity on-line estimation of hidden Markov model parameters

In this paper we propose and study low complexity algorithms for on-line estimation of hidden Markov model (HMM) parameters. The estimates approach the true model parameters as the measurement noise approaches zero, but otherwise give improved estimates, albeit with bias. On a nite data set in the high noise case, the bias may not be signi cantly more severe than for a higher complexity asymptotically optimal scheme. Our algorithms require O(N3) calculations per time instant, where N is the number of states. Previous algorithms based on earlier hidden Markov model signal processing methods, including the expectation-maximumisation (EM) algorithm require O(N4) calculations per time instant.

On-line estimation of Allan variance parameters

A new online method is presented for estimation of the angular random walk and rate random walk coefficients of IMU (inertial measurement unit) gyros and accelerometers. The online method proposes a state space model and proposes parameter estimators for quantities previously measured from off-line data techniques such as the Allan variance graph. Allan variance graphs have large off-line computational effort and data storage requirements. The technique proposed here requires no data storage and computational effort of O(100) calculations per data sample.

Item reputation-aware recommender systems

Recommender systems provide personalized advice for customers online based on their own preferences, while reputation systems generate a community advice on the quality of items on the Web. Both systems use users’ ratings to generate their output. In this paper, we propose to combine reputation models with recommender systems to enhance the accuracy of recommendations. The main contributions include two methods for merging two ranked item lists which are generated based on recommendation scores and reputation scores, respectively, and a personalized reputation method to generate item reputations based on users’ interests. The proposed merging methods can be applicable to any recommendation methods and reputation methods, i.e., they are independent from generating recommendation scores and reputation scores. The experiments we conducted showed that the proposed methods could enhance the accuracy of existing recommender systems.

An efficient tagging data interpretation and representation scheme for item recommendation

A tag-based item recommendation method generates an ordered list of items, likely interesting to a particular user, using the users past tagging behaviour. However, the users tagging behaviour varies in different tagging systems. A potential problem in generating quality recommendation is how to build user profiles, that interprets user behaviour to be effectively used, in recommendation models. Generally, the recommendation methods are made to work with specific types of user profiles, and may not work well with different datasets. In this paper, we investigate several tagging data interpretation and representation schemes that can lead to building an effective user profile. We discuss the various benefits a scheme brings to a recommendation method by highlighting the representative features of user tagging behaviours on a specific dataset. Empirical analysis shows that each interpretation scheme forms a distinct data representation which eventually affects the recommendation result. Results on various datasets show that an interpretation scheme should be selected based on the dominant usage in the tagging data (i.e. either higher amount of tags or higher amount of items present). The usage represents the characteristic of user tagging behaviour in the system. The results also demonstrate how the scheme is able to address the cold-start user problem.

Transport and health in China : recent collaborative research relating to vehicle factors and pollution

The nature of the transport system contributes to public health outcomes in a range of ways. The clearest contribution to public health is in the area of traffic crashes, because of their direct impact on individual death and disability and their direct costs to the health system. Other papers in this conference address these issues. This paper outlines some collaborative research between the Centre for Accident Research and Road Safety - Queensland (CARRS-Q) at QUT and Chinese researchers in areas that have indirect health impacts. Heavy vehicle dynamics: The integrity of the road surface influences crash risk, with ruts, pot-holes and other forms of road damage contributing to increased crash risks. The great majority of damage to the road surface from vehicles is caused by heavy trucks and buses, rather than cars or smaller vehicles. In some cases this damage is due to deliberate overloading, but in other cases it is due to vehicle suspension characteristics that lead to occasional high loads on particular wheels. Together with a visiting researcher and his colleagues, we have used both Queensland and Chinese data to model vehicle suspension systems that reduce the level of load, and hence the level of road damage and resulting crash risk(1-5). Toll worker exposure to vehicle emissions: The increasing construction of highways in China has also involved construction of a large number of toll roads. Tollbooth workers are potentially exposed to high levels of pollutants from vehicles, however the extent of this exposure and how it relates to standards for exposure are not well known. In a study led by a visiting researcher, we conducted a study to model these levels of exposure for a tollbooth in China(6). Noise pollution: The increasing presence of high speed roads in China has contributed to an increase in noise levels. In this collaborative study we modelled noise levels associated with a freeway widening near a university campus, and measures to reduce the noise(7). Along with these areas of research, there are many other areas of transport with health implications that are worthy of exploration. Traffic, noise and pollution contribute to a difficult environment for pedestrians, especially in an ageing society where there are health benefits to increasing physical activity. By building on collaborations such as those outlined, there is potential for a contribution to improved public health by addressing transport issues such as vehicle factors and pollution, and extending the research to other areas of travel activity. 1. Chen, Y., He, J., King, M., Chen, W. and Zhang, W. (2014). Stiffness-damping matching method of an ECAS system based on LQG control. Journal of Central South University, 21:439-446. DOI: 10.1007/s1177101419579 2. Chen, Y., He, J., King, M., Feng, Z. and Chang, W. (2013). Comparison of two suspension control strategies for multi-axle heavy truck. Journal of Central South University, 20(2): 550-562. 3. Chen, Y., He, J., King, M., Chen, W. and Zhang, W. (2013). Effect of driving conditions and suspension parameters on dynamic load-sharing of longitudinal-connected air suspensions. Science China Technological Sciences, 56(3): 666-676. DOI: 10.1007/s11431-012-5091-3 4. Chen, Y., He., J., King, M., Chen, W. and Zhang, W. (2013). Model development and dynamic load-sharing analysis of longitudinal-connected air suspensions. Strojniški Vestnik - Journal of Mechanical Engineering, 59(1):14-24. 5. Chen, Y., He, J., King, M., Liu, H. and Zhang, W. (2013). Dynamic load-sharing of longitudinal-connected air suspensions of a tri-axle semi-trailer. Proceedings of Transportation Research Board Annual Conference, Washington DC, 13-17 January 2013, paper no. 13-1117. 6. He, J., Qi, Z., Hang, W., King, M., and Zhao, C. (2011). Numerical evaluation of pollutant dispersion at a toll plaza based on system dynamics and Computational Fluid Dynamics models. Transportation Research Part C, 19(2011):510-520. 7. Zhang, C., He, J., Wang, Z., Yin, R. and King, M. (2013). Assessment of traffic noise level before and after freeway widening using traffic microsimulation and a refined classic noise prediction method. Proceedings of Transportation Research Board Annual Conference, Washington DC, 13-17 January 2013, paper no. 13-2016.

Retinal tissue thickness is reduced in diabetic peripheral neuropathy

Aim Retinal tissue integrity in relation to diabetic neuropathy is not known. The aim of this study was to investigate retinal tissue thickness in relation to diabetic peripheral neuropathy (DPN) with and without diabetic retinopathy (DR). Methods Full retinal thickness at the parafoveal and perifoveal macula and neuro-retinal thickness around the optic nerve head (ONH) and at the macula was examined using spectral domain optical coherence tomography. The eye on the hand-dominant side of 85 individuals with type 1 diabetes and 66 individuals with type 2 diabetes, with or without DR and DPN, were compared to the eyes (n=45) of age-matched non-diabetic controls. Diabetic neuropathy was defined as Neuropathy Disability Score (NDS) ≥3 on a scale of 0-10. A general linear model was used to examine the relationship between diabetic neuropathy and foveal, parafoveal and perifoveal retinal thickness and neuro-retinal thickness, in relation to DR status, age, gender, HbA1c levels and duration of diabetes. A p-value of <0.05 was considered statistically significant. Results Perifoveal retinal thickness is reduced with increasing severity of neuropathy, especially in the inferior hemisphere (p=0.004); this effect was not related to age (p=0.088). For every unit increase in NDS score, the inferior perifoveal retinal thickness reduced by 1.64 μm. Neuro-retinal thickness around the ONH decreased with increasing severity of neuropathy (p<0.014 for average and hemisphere thicknesses); for every unit increase in NDS, neuro-retinal thickness around the ONH reduced by 1.23 μm. Retinal thickness in the parafovea was increased in the absence of DR (p<0.017 for average and hemisphere thicknesses). Neuro-retinal thickness at the macula was inversely related to age alone (p<0.001). All retinal parameters, except the inferior perifovea, reduced with advancing age (p<0.007 for all). Conclusions Diabetic neuropathy is associated with changes in full retinal thickness and neuro-retinal layers. This may represent a second threat to vision integrity, in addition to the better-characterised retinopathy. This study provides new knowledge about the anatomical aspects of the retinal tissue in relation to neuropathy and retinopathy.

Porosity: Establishing the relationship between drying parameters and dried food quality [SpringerBriefs in Food, Health, and Nutrition]

The preservation technique of drying offers a significant increase in the shelf life of food materials, along with the modification of quality attributes due to simultaneous heat and mass transfer. Variations in porosity are just one of the microstructural changes that take place during the drying of most food materials. Some studies found that there may be a relationship between porosity and the properties of dried foods. However, no conclusive relationship has yet been established in the literature. This paper presents an overview of the factors that influence porosity, as well as the effects of porosity on dried food quality attributes. The effect of heat and mass transfer on porosity is also discussed along with porosity development in various drying methods. After an extensive review of the literature concerning the study of porosity, it emerges that a relationship between process parameters, food qualities, and sample properties can be established. Therefore, we propose a hypothesis of relationships between process parameters, product quality attributes, and porosity.

Refining user and item profiles based on multidimensional data for top-n item recommendation

In recommender systems based on multidimensional data, additional metadata provides algorithms with more information for better understanding the interaction between users and items. However, most of the profiling approaches in neighbourhood-based recommendation approaches for multidimensional data merely split or project the dimensional data and lack the consideration of latent interaction between the dimensions of the data. In this paper, we propose a novel user/item profiling approach for Collaborative Filtering (CF) item recommendation on multidimensional data. We further present incremental profiling method for updating the profiles. For item recommendation, we seek to delve into different types of relations in data to understand the interaction between users and items more fully, and propose three multidimensional CF recommendation approaches for top-N item recommendations based on the proposed user/item profiles. The proposed multidimensional CF approaches are capable of incorporating not only localized relations of user-user and/or item-item neighbourhoods but also latent interaction between all dimensions of the data. Experimental results show significant improvements in terms of recommendation accuracy.

Text mining for screening efficiency? Testing within a Cochrane public health review

Background The requirement for dual screening of titles and abstracts to select papers to examine in full text can create a huge workload, not least when the topic is complex and a broad search strategy is required, resulting in a large number of results. An automated system to reduce this burden, while still assuring high accuracy, has the potential to provide huge efficiency savings within the review process. Objectives To undertake a direct comparison of manual screening with a semi‐automated process (priority screening) using a machine classifier. The research is being carried out as part of the current update of a population‐level public health review. Methods Authors have hand selected studies for the review update, in duplicate, using the standard Cochrane Handbook methodology. A retrospective analysis, simulating a quasi‐‘active learning’ process (whereby a classifier is repeatedly trained based on ‘manually’ labelled data) will be completed, using different starting parameters. Tests will be carried out to see how far different training sets, and the size of the training set, affect the classification performance; i.e. what percentage of papers would need to be manually screened to locate 100% of those papers included as a result of the traditional manual method. Results From a search retrieval set of 9555 papers, authors excluded 9494 papers at title/abstract and 52 at full text, leaving 9 papers for inclusion in the review update. The ability of the machine classifier to reduce the percentage of papers that need to be manually screened to identify all the included studies, under different training conditions, will be reported. Conclusions The findings of this study will be presented along with an estimate of any efficiency gains for the author team if the screening process can be semi‐automated using text mining methodology, along with a discussion of the implications for text mining in screening papers within complex health reviews.