Search

Abstract An excellent cardinality estimation can make the query optimiser produce a good execution plan. Although there are some studies on cardinality estimation, the prediction results of existing cardinality estimators are inaccurate and the query efficiency cannot be guaranteed as well. In particular, they are difficult to accurately obtain the complex relationships between multiple tables in complex database systems. When dealing with complex queries, the existing cardinality estimators cannot achieve good results. In this study, a novel cardinality estimator is proposed. It uses the core techniques with the BiLSTM network structure and adds the attention mechanism. First, the columns involved in the query statements in the training set are sampled and compressed into bitmaps. Then, the Word2vec model is used to embed the word vectors about the query statements. Finally, the BiLSTM network and attention mechanism are employed to deal with word vectors. The proposed model takes into consideration not only the correlation between tables but also the processing of complex predicates. Extensive experiments and the evaluation of BiLSTM‐Attention Cardinality Estimator (BACE) on the IMDB datasets are conducted. The results show that the deep learning model can significantly improve the quality of cardinality estimation, which is a vital role in query optimisation for complex databases.

Abstract With the development of location‐based services and Big data technology, vehicle map matching techniques are growing rapidly, which is the fundamental techniques in the study of exploring global positioning system (GPS) data. The pre‐processed GPS data can provide the guarantee of high‐quality data for the research of mining passenger’s points of interest and urban computing services. The existing surveys mainly focus on map‐matching algorithms, but there are few descriptions on the key phases of the acquisition of sampling data, floating car and road data preprocessing in vehicle map matching systems. To address these limitations, the contribution of this survey on map matching techniques lies in the following aspects: (i) the background knowledge, function and system framework of vehicle map matching techniques; (ii) description of floating car data and road network structure to understand the detailed phase of map matching; (iii) data preprocessing rules, specific methodologies, and significance of floating car and road data; (iv) map matching algorithms are classified by the sampling frequency and data information. The authors give the introduction of open‐source GPS sampling data sets, and the evaluation measurements of map‐matching approaches; (v) the suggestions on data preprocessing and map matching algorithms in the future work.

The problem of classifying imbalanced data is one of the active research directions in machine learning and bioinformatics. The imbalance of data will greatly degrade the accuracy of classifiers. Good oversampling methods can improve the diversity and validity of new samples, which cannot only solve the imbalance problem of sample data, but also greatly improve the classification accuracy. In this study, we propose the trigonal barycenter theory and a semi-supervised oversampling method, called SOTB (Semi-supervised Oversampling method based on Trigonal Barycenter theory). SOBT works to: (1) construct the non-intersecting triangles based on Mahalanobis distance; (2) combine the semi-supervised sampling method with trigonal barycenter theory to oversample the positive samples, which can cope with the data imbalance problem without affecting the quality of data. Lastly, extensive experiments were conducted to verify the effectiveness of the proposed method. The results demonstrate that SOTB can improve the validity, diversity and rationality on the distribution of the newly generated samples as well as alleviate the phenomena of over-fitting which is popular in existing oversampling approaches. In particular, when compared with the state-of-the-art oversampling methods, the results show SOTB can achieve the best classification performance.

Natural scene text detection has attracted much attention in the research field of computer vision, and it has been widely used in many applications, i.e., unmanned driving, robot sensing. Some methods have been proposed for horizontal and oriented text detection, but detecting irregular shapes and highly varying orientated text is still a challenging problem. To tackle this problem, we propose a robust arbitrary shape text detection method called Soft Dilated network (SDnet). The proposed method has two essential steps: (1) feature extraction by backbone; (2) post-processing approach to generate elaborated polygons or boundaries. In particular, the backbone is based on soft attention mechanism and dilated convolution. The soft attention mechanism learns and obtains importance feature from each feature channel, and dilated convolution can effectively aggregate multi-scale contextual information without losing the resolution, and enhance the robust of the network model. The proposed method can accurately detect curve text and discriminate text and non-text areas in an efficient fashion. In addition, Jaccard coefficient is used as loss function to promote the post-processing capability of detecting sparse-arranged and arbitrary shape text. Based on the aforementioned technique, the proposed method an effectively handle the problem of sparse arranged arbitrary natural scene text detection. Experiments were conducted on three benchmark datasets: curved text dataset CTW1500, Total-Text and oriented dataset ICDAR2015, and the results show that when compared with the state-of-the-art text detection methods, the proposed method is more robust and it can find smaller text blocks in the image due to the Loss Function calculation with Jaccard coefficient. Furthermore, we performed multiple sets of ablation experiments, verify the effectiveness of the propose method.

Social networks are extensively exploited by third-party consumers such as researchers and advertisers to understand user characteristics and behaviors. In general, before network data is published, sensitive relationships should be anonymized to prevent the compromise of individual privacy. To quantify the guarantee level of privacy-preserving mechanisms and mitigate users' privacy concerns, numerous studies concerning network data de-anonymization have been carried out. However, most existing studies focus on single-view data, and privacy protection for multi-view data that is ubiquitous in the era of big data has not been yet extensively explored. In this study, we are interested in answering the following question: Are the traditional privacy protection methods still valid for the anonymization of multi-view data? In this study, we propose a Multi-View Low-Rank Coding (MVLRC) based network data de-anonymization framework to assess the vulnerability of privacy protection techniques by accurately reconstructing a large portion of the original data. Specifically, the framework assumes that in principle, the target and auxiliary networks have common structural patterns, and they can be modeled together to infer the hidden structure of the target network. The essential components of our work include the following: (1) a robust network representation model for structural pattern learning; (2) the network representation based multi-view modeling of target network and auxiliary network; (3) the inference of the anonymized links via target network reconstruction. Experimental results on synthetic networks and three real-world networks demonstrate that auxiliary networks can be utilized by malicious adversaries for privacy inference attacks. Thus, the privacy protection of multi-view network data needs more sophisticated anonymization techniques.

Abstract Background Imbalanced datasets are commonly encountered in bioinformatics classification problems, that is, the number of negative samples is much larger than that of positive samples. Particularly, the data imbalance phenomena will make us underestimate the performance of the minority class of positive samples. Therefore, how to balance the bioinformatic data becomes a very challenging and difficult problem. Results In this study, we propose a new data sampling approach, called pseudo-negative sampling, which can be effectively applied to handle the case that: negative samples greatly dominate positive samples. Specifically, we design a supervised learning method based on a max-relevance min-redundancy criterion beyond Pearson correlation coefficient (MMPCC), which is used to choose pseudo-negative samples from the negative samples and view them as positive samples. In addition, MMPCC uses an incremental searching technique to select optimal pseudo-negative samples to reduce the computation cost. Consequently, the discovered pseudo-negative samples have strong relevance to positive samples and less redundancy to negative ones. Conclusions To validate the performance of our method, we conduct experiments base on four UCI datasets and three real bioinformatics datasets. According to the experimental results, we clearly observe the performance of MMPCC is better than other sampling methods in terms of Sensitivity, Specificity, Accuracy and the Mathew’s Correlation Coefficient. This reveals that the pseudo-negative samples are particularly helpful to solve the imbalance dataset problem. Moreover, the gain of Sensitivity from the minority samples with pseudo-negative samples grows with the improvement of prediction accuracy on all dataset.

Graphs have been proved to be a useful mathematical representation for a broad variety of real-world complex systems, and the structure prediction on graphs refers to estimating the potential relationship between the objects from the observed structures, being fundamental in many data analysis applications, such as network alignment, network reconstruction, and link prediction. Accordingly, in data publishing, it is necessary to regulate the structural predictability of graphs against inference attack to protect the sensitive information of the data generators. In contrast to the existing works about graph structure perturbation for node ranking, information diffusion, and so on, the structural predictability optimization problem, i.e., reducing the accuracy of sensitive relationships inference in graphs, has not been extensively studied. This paper presents an active learning algorithm that selects the most representative links to be perturbed, thus regulating the structural predictability of graphs, that is, removing as few as possible links to undermine the regularity level of graphs, which forms the foundation of inference attack methods. Specifically, with the assumption that the substructure with higher regularity level contains more regular equivalence components and has more equivalent paths supplied for the random walk processes, random walk-based link importance measuring algorithm is proposed to identify the representative links. The structural regularity metric, measuring the structural predictability of graphs, is also introduced to guide the link perturbation for structural predictability optimization. The extensive experiments on artificial and real-world data sets demonstrate the effectiveness of the proposed structural predictability optimization method. Specifically, the method can learn the role of links accurately in term of graph organization, and the performance of structure inference on graphs can be deteriorated effectively by representative link-based perturbation.

With the development of network services and location-based systems, many mobile applications begin to use users’ geographical location to provide better services. In terms of social networks, geographical location is actively shared by users. In some applications with recommendation services, before the geographical location recommendation is provided, the authors have to obtain user's permission. This kind of social network integrated with geographical location information is called location-based social networks (abbreviate for LBSNs). In the LBSN, each user has location information when he or she checked in hotels or feature spots. Based on this information, they can identify user's trajectory of movement behaviour and activity patterns. In general, if there is friendship between two users, their trajectories in reality are likely to be similar. In this study, according to user's geographical location information over a period of time, they explore whether there exists friendly relationship between two users based on trajectory similarity and the structure theory of graphs. In particular, they propose a new factor function and a factor graph model based on user's geographical location to predict the friendship between two users in the real LBSN.

The existing continuous nearest neighbor query algorithms of moving objects in road networks do not consider any traffic rule and assume that the speed of moving objects is constant and the topology of road networks never change. However, in real road networks, the object’s speed and the road network’s structure change frequently Hence, these would make the existing methods ineffective when applying to the real-world road network environment To overcome the aforementioned disadvantages, we propose a Data Modeling approach of Road Networks with traffic rules (called DMRNR) and design a novel Expanding Tree-based Continuous k Nearest Neighbors algorithm (abbreviate for ExTCKNN) that can be well adopted to the actual road network environment. The algorithm consists of three steps: 1) it obtains the query results to store using DMRNR in the initial phase; 2) it maintains the data model of road networks by monitoring the real-time change information; and 3) the results are generated according to the submitted query with the updated data model and the latest state of moving objects The merit of the proposed algorithm lies in that it queries the nearest neighbors by taking the movements of the moving object and the variety of the road networks into consideration Extensive experiments are conducted and the experimental results demonstrate a significant improvement of the proposed method when compared with conventional solutions.

To effectively score pages with uncertainty in web social networks, we first proposed a new concept called transition probability matrix and formally defined the uncertainty in web social networks. Second, we proposed a hybrid page scoring algorithm, called WebScore, based on the PageRank algorithm and three centrality measures including degree, betweenness, and closeness. Particularly,WebScore takes into a full consideration of the uncertainty of web social networks by computing the transition probability from one page to another. The basic idea ofWebScore is to: (1) integrate uncertainty into PageRank in order to accurately rank pages, and (2) apply the centrality measures to calculate the importance of pages in web social networks. In order to verify the performance of WebScore, we developed a web social network analysis system which can partition web pages into distinct groups and score them in an effective fashion. Finally, we conducted extensive experiments on real data and the results show that WebScore is effective at scoring uncertain pages with less time deficiency than PageRank and centrality measures based page scoring algorithms.

This paper proposes the concept of intervention rule which tries to reveal the interventional relationship between elements in a system in the following three aspects. (1) Casual relationship. Intervention rule shows which element is the cause and which element is the consequence. (2) Quantitative relationship: Intervention rule shows the quantitative intensity of how the change of the causal element interferes with the change of the consequential element. (3) Multi-scale intervention relationship. Intervention rule shows the intervention at different decomposition scale of the original system, since sub system may exhibit different mechanism from the original system. This paper first introduces a general intervention rule framework, and then transforms the framework into concrete intervention rules for complex network data and time series data. Then, it proposes two algorithms to mine the intervention rules from the two different systems. Finally, the experimental results show that multi-scale intervention rules do exist in real dataset. And the intervention intensity of each sub graph and sub series are always 4 or 5 time larger than intervention intensity of the original data.

The trajectory pattern mining problem has recently attracted much attention due to the rapid development of location-acquisition technologies, and parallel computing essentially provides an alternative method for handling this problem. This study precisely addresses the problem of parallel mining of trajectory sequential patterns based on the newly proposed concepts with regard to trajectory pattern mining. We propose an efficient and effective parallel sequential patterns mining (plute) algorithm that includes three essential techniques: prefix projection, data parallel formulation, and task parallel formulation. Firstly, the prefix projection technique is used to decompose the search space as well as greatly reduce the candidate trajectory sequences. Secondly, the data parallel formulation decomposes the computations associated with counting the support of trajectory patterns. Thirdly, the task parallel formulation employs the MapReduce programming model to assign the computations across a set of machines in a scalable and easy-to-use fashion. Based on the properties of parallel trajectory sequences, item pruning and sequence pruning strategies are applied to further prune the candidate sequences. Extensive experiments are conducted to evaluate the performance of plute in terms of parallel computing time and communication cost among processors. Experimental results show that plute outperforms the previously proposed parallel mining strategy (PartSpan) in mining massive trajectory data.