Similarity from Multi-Dimensional Scaling: Solving the Accuracy and Diversity Dilemma in Information Filtering

Recommender systems are designed to assist individual users to navigate through the rapidly growing amount of information. One of the most successful recommendation techniques is the collaborative filtering, which has been extensively investigated and has already found wide applications in e-commerce. One of challenges in this algorithm is how to accurately quantify the similarities of user pairs and item pairs. In this paper, we employ the multidimensional scaling (MDS) method to measure the similarities between nodes in user-item bipartite networks. The MDS method can extract the essential similarity information from the networks by smoothing out noise, which provides a graphical display of the structure of the networks. With the similarity measured from MDS, we find that the item-based collaborative filtering algorithm can outperform the diffusion-based recommendation algorithms. Moreover, we show that this method tends to recommend unpopular items and increase the global diversification of the networks in long term.     

Information Filtering in Sparse Online Systems: Recommendation via Semi-Local Diffusion

With the rapid growth of the Internet and overwhelming amount of information and choices that people are confronted with, recommender systems have been developed to effectively support users’ decision-making process in the online systems. However, many recommendation algorithms suffer from the data sparsity problem, i.e. the user-object bipartite networks are so sparse that algorithms cannot accurately recommend objects for users. This data sparsity problem makes many well-known recommendation algorithms perform poorly. To solve the problem, we propose a recommendation algorithm based on the semi-local diffusion process on the user-object bipartite network. The simulation results on two sparse datasets, Amazon and Bookcross, show that our method significantly outperforms the state-of-the-art methods especially for those small-degree users. Two personalized semi-local diffusion methods are proposed which further improve the recommendation accuracy. Finally, our work indicates that sparse online systems are essentially different from the dense online systems, so it is necessary to reexamine former algorithms and conclusions based on dense data in sparse systems.     

AUI&GIV: Recommendation with Asymmetric User Influence and Global Importance Value

The user-based collaborative filtering (CF) algorithm is one of the most popular approaches for making recommendation. Despite its success, the traditional user-based CF algorithm suffers one serious problem that it only measures the influence between two users based on their symmetric similarities calculated by their consumption histories. It means that, for a pair of users, the influences on each other are the same, which however may not be true. Intuitively, an expert may have an impact on a novice user but a novice user may not affect an expert at all. Besides, each user may possess a global importance factor that affects his/her influence to the remaining users. To this end, in this paper, we propose an asymmetric user influence model to measure the directed influence between two users and adopt the PageRank algorithm to calculate the global importance value of each user. And then the directed influence values and the global importance values are integrated to deduce the final influence values between two users. Finally, we use the final influence values to improve the performance of the traditional user-based CF algorithm. Extensive experiments have been conducted, the results of which have confirmed that both the asymmetric user influence model and global importance value play key roles in improving recommendation accuracy, and hence the proposed method significantly outperforms the existing recommendation algorithms, in particular the user-based CF algorithm on the datasets of high rating density.     

The Power of Ground User in Recommender Systems

Accuracy and diversity are two important aspects to evaluate the performance of recommender systems. Two diffusion-based methods were proposed respectively inspired by the mass diffusion (MD) and heat conduction (HC) processes on networks. It has been pointed out that MD has high recommendation accuracy yet low diversity, while HC succeeds in seeking out novel or niche items but with relatively low accuracy. The accuracy-diversity dilemma is a long-term challenge in recommender systems. To solve this problem, we introduced a background temperature by adding a ground user who connects to all the items in the user-item bipartite network. Performing the HC algorithm on the network with ground user (GHC), it showed that the accuracy can be largely improved while keeping the diversity. Furthermore, we proposed a weighted form of the ground user (WGHC) by assigning some weights to the newly added links between the ground user and the items. By turning the weight as a free parameter, an optimal value subject to the highest accuracy is obtained. Experimental results on three benchmark data sets showed that the WGHC outperforms the state-of-the-art method MD for both accuracy and diversity.     

Measuring User Similarity Using Electric Circuit Analysis: Application to Collaborative Filtering

We propose a new technique of measuring user similarity in collaborative filtering using electric circuit analysis. Electric circuit analysis is used to measure the potential differences between nodes on an electric circuit. In this paper, by applying this method to transaction networks comprising users and items, i.e., user–item matrix, and by using the full information about the relationship structure of users in the perspective of item adoption, we overcome the limitations of one-to-one similarity calculation approach, such as the Pearson correlation, Tanimoto coefficient, and Hamming distance, in collaborative filtering. We found that electric circuit analysis can be successfully incorporated into recommender systems and has the potential to significantly enhance predictability, especially when combined with user-based collaborative filtering. We also propose four types of hybrid algorithms that combine the Pearson correlation method and electric circuit analysis. One of the algorithms exceeds the performance of the traditional collaborative filtering by 37.5% at most. This work opens new opportunities for interdisciplinary research between physics and computer science and the development of new recommendation systems     

Item Anomaly Detection Based on Dynamic Partition for Time Series in Recommender Systems

In recent years, recommender systems have become an effective method to process information overload. However, recommendation technology still suffers from many problems. One of the problems is shilling attacks-attackers inject spam user profiles to disturb the list of recommendation items. There are two characteristics of all types of shilling attacks: 1) Item abnormality: The rating of target items is always maximum or minimum; and 2) Attack promptness: It takes only a very short period time to inject attack profiles. Some papers have proposed item anomaly detection methods based on these two characteristics, but their detection rate, false alarm rate, and universality need to be further improved. To solve these problems, this paper proposes an item anomaly detection method based on dynamic partitioning for time series. This method first dynamically partitions item-rating time series based on important points. Then, we use chi square distribution (χ²) to detect abnormal intervals. The experimental results on MovieLens 100K and 1M indicate that this approach has a high detection rate and a low false alarm rate and is stable toward different attack models and filler sizes.     

Optimizing Online Social Networks for Information Propagation

Online users nowadays are facing serious information overload problem. In recent years, recommender systems have been widely studied to help people find relevant information. Adaptive social recommendation is one of these systems in which the connections in the online social networks are optimized for the information propagation so that users can receive interesting news or stories from their leaders. Validation of such adaptive social recommendation methods in the literature assumes uniform distribution of users'' activity frequency. In this paper, our empirical analysis shows that the distribution of online users'' activity is actually heterogenous. Accordingly, we propose a more realistic multi-agent model in which users'' activity frequency are drawn from a power-law distribution. We find that previous social recommendation methods lead to serious delay of information propagation since many users are connected to inactive leaders. To solve this problem, we design a new similarity measure which takes into account users'' activity frequencies. With this similarity measure, the average delay is significantly shortened and the recommendation accuracy is largely improved.     

Information Filtering via Heterogeneous Diffusion in Online Bipartite Networks

The rapid expansion of Internet brings us overwhelming online information, which is impossible for an individual to go through all of it. Therefore, recommender systems were created to help people dig through this abundance of information. In networks composed by users and objects, recommender algorithms based on diffusion have been proven to be one of the best performing methods. Previous works considered the diffusion process from user to object, and from object to user to be equivalent. We show in this work that it is not the case and we improve the quality of the recommendation by taking into account the asymmetrical nature of this process. We apply this idea to modify the state-of-the-art recommendation methods. The simulation results show that the new methods can outperform these existing methods in both recommendation accuracy and diversity. Finally, this modification is checked to be able to improve the recommendation in a realistic case.     

Promoting Cold-Start Items in Recommender Systems

As one of the major challenges, cold-start problem plagues nearly all recommender systems. In particular, new items will be overlooked, impeding the development of new products online. Given limited resources, how to utilize the knowledge of recommender systems and design efficient marketing strategy for new items is extremely important. In this paper, we convert this ticklish issue into a clear mathematical problem based on a bipartite network representation. Under the most widely used algorithm in real e-commerce recommender systems, the so-called item-based collaborative filtering, we show that to simply push new items to active users is not a good strategy. Interestingly, experiments on real recommender systems indicate that to connect new items with some less active users will statistically yield better performance, namely, these new items will have more chance to appear in other users'' recommendation lists. Further analysis suggests that the disassortative nature of recommender systems contributes to such observation. In a word, getting in-depth understanding on recommender systems could pave the way for the owners to popularize their cold-start products with low costs.     

Mining-based associative image filtering using harmonic mean

With the development of IT convergence technologies, users can now more easily access useful information. These days, diverse and far-reaching information is being rapidly produced and distributed instantly in digitized format. Studies are continuously seeking to develop more efficient methods of delivering information to a greater number of users. Image filtering, which extracts features of interest from images, was developed to address the weakness of collaborative filtering, which is limited to superficial data analysis. However, image filtering has its own weakness of requiring complicated calculations to obtain the similarity between images. In this study, to resolve these problems, we propose associative image filtering based on the mining method utilizing the harmonic mean. Using data mining’s Apriori algorithm, this study investigated the association among preferred images from an associative image group and obtained a prediction based on user preference mean. In so doing, we observed a positive relationship between the various image preferences and the various distances between images’ color histograms. Preference mean was calculated based on the arithmetic mean, geometric mean, and harmonic mean. We found through performance analysis that the harmonic mean had the highest accuracy. In associative image filtering, we used the harmonic mean in order to anticipate preferences. In testing accuracy with MAE utilizing the proposed method, this study demonstrated an improvement of approximately 12 % on average compared to previous collaborative image filtering.     

Using Dynamic Multi-Task Non-Negative Matrix Factorization to Detect the Evolution of User Preferences in Collaborative Filtering

Predicting what items will be selected by a target user in the future is an important function for recommendation systems. Matrix factorization techniques have been shown to achieve good performance on temporal rating-type data, but little is known about temporal item selection data. In this paper, we developed a unified model that combines Multi-task Non-negative Matrix Factorization and Linear Dynamical Systems to capture the evolution of user preferences. Specifically, user and item features are projected into latent factor space by factoring co-occurrence matrices into a common basis item-factor matrix and multiple factor-user matrices. Moreover, we represented both within and between relationships of multiple factor-user matrices using a state transition matrix to capture the changes in user preferences over time. The experiments show that our proposed algorithm outperforms the other algorithms on two real datasets, which were extracted from Netflix movies and Last.fm music. Furthermore, our model provides a novel dynamic topic model for tracking the evolution of the behavior of a user over time.     

Incorporating service and user information and latent features to predict QoS for selecting and recommending cloud service compositions

The rapid growth of published cloud services in the Internet makes the service selection and recommendation a challenging task for both users and service providers. In cloud environments, software re services collaborate with other complementary services to provide complete solutions to end users. The service selection is performed based on QoS requirements submitted by end users. Software providers alone cannot guarantee users’ QoS requirements. These requirements must be end-to-end, representing all collaborating services in a cloud solution. In this paper, we propose a prediction model to compute end-to-end QoS values for vertically composed services which are composed of three types of cloud services: software (SaaS), infrastructure (IaaS) and data (DaaS) services. These values can be used during the service selection and recommendation process. Our model exploits historical QoS values and cloud service and user information to predict unknown end-to-end QoS values of composite services. The experiments demonstrate that our proposed model outperforms other prediction models in terms of the prediction accuracy. We also study the impact of different parameters on the prediction results. In the experiments, we used real cloud services’ QoS data collected using our developed QoS monitoring and collecting system.     

A Network Centrality Method for the Rating Problem

We propose a new method for aggregating the information of multiple users rating multiple items. Our approach is based on the network relations induced between items by the rating activity of the users. Our method correlates better than the simple average with respect to the original rankings of the users, and besides, it is computationally more efficient than other methods proposed in the literature. Moreover, our method is able to discount the information that would be obtained adding to the system additional users with a systematically biased rating activity.     

A context-aware personalized resource recommendation for pervasive learning

As it is difficult for learners to discover and obtain the most appropriate resources from massive education resources according to traditional keyword searching method, the context-aware based resource recommendation service becomes a significant part of pervasive learning environments. At present, recommendation mechanisms are widely used in e-commerce field, where content-based or collaborative-based filter strategies are usually considered separately. However, in these existing recommendation mechanisms, the dynamic interests and preference of learners, the access pattern and the other attributes of pervasive learning environments (such as multi-modes connecting and resources distribution) are always neglected. Thus, these mechanisms can not effectively reflect learners’ actual preference and can not adapt to pervasive learning environments perfectly. To address these problems, a context-aware resource recommendation model and relevant recommendation algorithm for pervasive learning environments are proposed. Therein, with taking into account the relevant contextual information, the calculation of relevant degree between learners and resources can be divided into two main parts: logic-based RRD (resource relevant degree) and situation-based RRD. In the first part, content-based and collaborative-based recommendation mechanisms are combined together, where the individual preference tree (IPT) is introduced to take into account the multi-dimensional attributes of resources, learners’ rating matrix and the energy of access preference. Meanwhile, the learners’ historical sequential patterns of resource accessing are also considered to further improve the accuracy of recommendation. In the second part, in order to enhance the validation of recommendation, the connecting type relevance and time satisfaction degree are calculated according to other relevant contexts. Then, the candidate resources can be filtered and sorted via combining these two parts to generate (Top-N) recommendation results. The simulations show that our newly proposed method outperforms other state of-the-art algorithms on traditional and newly presented metrics and it may also be more suitable for pervasive learning environments. Finally, a prototype system is implemented based on SEU-ESP to demonstrate the relevant recommendation process further.     

Imputing missing data in plant traits: A guide to improve gap-filling

Aim

Globally distributed plant trait data are increasingly used to understand relationships between biodiversity and ecosystem processes. However, global trait databases are sparse because they are compiled from many, mostly small databases. This sparsity in both trait space completeness and geographical distribution limits the potential for both multivariate and global analyses. Thus, ‘gap-filling’ approaches are often used to impute missing trait data. Recent methods, like Bayesian hierarchical probabilistic matrix factorization (BHPMF), can impute large and sparse data sets using side information. We investigate whether BHPMF imputation leads to biases in trait space and identify aspects influencing bias to provide guidance for its usage.

Innovation

We use a fully observed trait data set from which entries are randomly removed, along with extensive but sparse additional data. We use BHPMF for imputation and evaluate bias by: (1) accuracy (residuals, RMSE, trait means), (2) correlations (bi- and multivariate) and (3) taxonomic and functional clustering (valuewise, uni- and multivariate). BHPMF preserves general patterns of trait distributions but induces taxonomic clustering. Data set–external trait data had little effect on induced taxonomic clustering and stabilized trait–trait correlations.

Main Conclusions

Our study extends the criteria for the evaluation of gap-filling beyond RMSE, providing insight into statistical data structure and allowing better informed use of imputed trait data, with improved practice for imputation. We expect our findings to be valuable beyond applications in plant ecology, for any study using hierarchical side information for imputation.     

Restriction data from chloroplast DNA for phylogenetic reconstruction: Is there only one accurate way of scoring?

Information from the same restriction analysis of chloroplast DNA of 33 taxa ofRubiaceae was scored in four different ways, two of which were based on fragments, and two on restriction sites, and they were subsequently analysed with Wagner parsimony. The methods resulted in different phylogenetic trees. The inherent differences between the methods relate to the amount of non-homologous characters and dependent characters, but none of the methods will systematically bias the resulting cladograms. The fragment analyses are much less time-consuming, but probably less accurate, than the site analyses. The choice of method is dependent on a trade-off between accuracy and resources (time). One important recommendation is made: all phylogenetic analyses of chloroplast DNA data should be accompanied by a data matrix and contain information on how the matrix was compiled.     

A Novel Semi-Supervised Methodology for Extracting Tumor Type-Specific MRS Sources in Human Brain Data

Background

The clinical investigation of human brain tumors often starts with a non-invasive imaging study, providing information about the tumor extent and location, but little insight into the biochemistry of the analyzed tissue. Magnetic Resonance Spectroscopy can complement imaging by supplying a metabolic fingerprint of the tissue. This study analyzes single-voxel magnetic resonance spectra, which represent signal information in the frequency domain. Given that a single voxel may contain a heterogeneous mix of tissues, signal source identification is a relevant challenge for the problem of tumor type classification from the spectroscopic signal.

Methodology/Principal Findings

Non-negative matrix factorization techniques have recently shown their potential for the identification of meaningful sources from brain tissue spectroscopy data. In this study, we use a convex variant of these methods that is capable of handling negatively-valued data and generating sources that can be interpreted as tumor class prototypes. A novel approach to convex non-negative matrix factorization is proposed, in which prior knowledge about class information is utilized in model optimization. Class-specific information is integrated into this semi-supervised process by setting the metric of a latent variable space where the matrix factorization is carried out. The reported experimental study comprises 196 cases from different tumor types drawn from two international, multi-center databases. The results indicate that the proposed approach outperforms a purely unsupervised process by achieving near perfect correlation of the extracted sources with the mean spectra of the tumor types. It also improves tissue type classification.

Conclusions/Significance

We show that source extraction by unsupervised matrix factorization benefits from the integration of the available class information, so operating in a semi-supervised learning manner, for discriminative source identification and brain tumor labeling from single-voxel spectroscopy data. We are confident that the proposed methodology has wider applicability for biomedical signal processing.     

Chlorophyll Fluorescence Detected Passively by Difference Reflectance Spectra of Wheat （Triticum aestivum L.） Leaf

The chlorophyll fluorescence （CF） signature emitted from vegetation provides an abundance of information regarding photosynthetics activity and has been used as a powerful tool to obtain physiological information of plant leaves in a non-invasive manner. CF is difficult to quantify because the CF signal is obscured by reflected light. In the present study, the apparent reflectance spectra of wheat （Triticum aestivum L.） leaves were measured under illuminations with and without filtering by three specially designed long-wave pass edge filters; the cut-off wavelengths of the three filters were 653.8, 678.2, and 694. l nm at 50% of maximum transmittance. The CF spectra could be derived as the reflectance difference spectra of the leaves under illuminations with and without the long wave pass edge filters. The ratio of the reflectance difference at 685 and 740 nm （Dif685/Dif740） was linear correlated with the CF parameters （maximal photochemical efficiency Fv/Fm, and the yield of quantum efficiency） measured by the modulated fluorometer. In addition, the ratio reflected the water stress status of the wheat leaf, which was very high when water deficiency was serious. This method provides a new approach for detecting CF and the physiological state of crops.     

Uncovering Community Structures with Initialized Bayesian Nonnegative Matrix Factorization

Uncovering community structures is important for understanding networks. Currently, several nonnegative matrix factorization algorithms have been proposed for discovering community structure in complex networks. However, these algorithms exhibit some drawbacks, such as unstable results and inefficient running times. In view of the problems, a novel approach that utilizes an initialized Bayesian nonnegative matrix factorization model for determining community membership is proposed. First, based on singular value decomposition, we obtain simple initialized matrix factorizations from approximate decompositions of the complex network’s adjacency matrix. Then, within a few iterations, the final matrix factorizations are achieved by the Bayesian nonnegative matrix factorization method with the initialized matrix factorizations. Thus, the network’s community structure can be determined by judging the classification of nodes with a final matrix factor. Experimental results show that the proposed method is highly accurate and offers competitive performance to that of the state-of-the-art methods even though it is not designed for the purpose of modularity maximization.     

Improving molecular cancer class discovery through sparse non-negative matrix factorization

MOTIVATION: Identifying different cancer classes or subclasses with similar morphological appearances presents a challenging problem and has important implication in cancer diagnosis and treatment. Clustering based on gene-expression data has been shown to be a powerful method in cancer class discovery. Non-negative matrix factorization is one such method and was shown to be advantageous over other clustering techniques, such as hierarchical clustering or self-organizing maps. In this paper, we investigate the benefit of explicitly enforcing sparseness in the factorization process. RESULTS: We report an improved unsupervised method for cancer classification by the use of gene-expression profile via sparse non-negative matrix factorization. We demonstrate the improvement by direct comparison with classic non-negative matrix factorization on the three well-studied datasets. In addition, we illustrate how to identify a small subset of co-expressed genes that may be directly involved in cancer.