A mobile social network system is generally complex that it comprises several views, such as data, function, structure, behavior and so on. There are two kinds of approaches to design these views. The multiple diagrams approach for mobile social network systems respectively chooses a distinct diagram for each view. The single diagram approach for mobile social network systems, instead of choosing several separated diagrams, utilizes only a single diagram. We propose the structure-behavior coalescence design method for mobile social network systems based on the single diagram approach to prevent the inter-diagram design inconsistency problems.

Unified Modeling Language (UML) possesses the model multiplicity problems because it does not integrate the systems structure with the systems behavior. To overcome this shortcoming, we propose an approach to enrich UML from model multiplicity to model singularity by creating its kernel model with the structure-behavior coalescence process algebra. The UML structure and behavior models will be derived from this kernel model. The characteristics of the single kernel model prevents the UML from the model multiplicity problem.

The popularity and development of mobile devices and mobile apps have dramatically changed human life. Due to the tremendous and still rapidly growing number of mobile apps, helping users find apps that satisfy their demands remains a difficult task. To address this problem, we propose a personalized mobile app recommender system based on the textual data of user reviews on the app store. Topic modeling techniques are applied to extract hidden topics of user reviews, and the probability distributions of the topics are utilized to represent the features of the apps. Then, the user profile is constructed based on the user's installed apps to capture user preferences. Both the topic distributions of the apps and user preferences are taken into account to produce recommendation scores to generate recommendation lists for target users. We crawl real-world data sets from app stores to evaluate the performance. The experimental results show that user reviews are effective for deriving the features of apps, and the proposed user-review-based app recommender system improves the performance of existing approaches. We conclude that the user reviews on the app store effectively represent the features of apps and play a significant role in personalized app recommender systems.

The financial media news reports have helped market participants to know about the strength and weakness of corporate future development, they still encounter a new challenge when analyzing businesses, because of their velocity, volume, variety, and veracity (4V). In comparison with the numerical ratios derived from financial reports, we believe textual information derived from financial media news reports provides much more essential and valuable cues pertaining to the existence of corporate financial difficulties, because the preparations of financial reports have to satisfy some selective accounting principle or estimation method determination. How to effectively and efficiently handle a large amount of textual content is an important in this explosive information period of big data analytics. To deal with the aforementioned challenge, this study implements a latent topic modelling technique and text mining technique to construct the “Intensity of Performance-Corpus Index” (IPCI), which can compress tremendous amounts of textual information into more manageable essential synthesized parts and then use them to forecast corporate operating performance. To our knowledge, current research has not yet constructed a forecasting model for corporate operating performance that simultaneously integrates numerical and textual information content. To fill the gap in the literature, this study combines numerical and textual information contents and uses them to construct a novel forecasting mechanism. With superior forecasting performance that is supported by real cases, the introduced mechanism can assist both market participants (such as bankers, investors, and creditors) and public sectors (such as Financial Supervisory Commission and Ministry of Economic Affairs) in allocating economic resources to form their own judgement as well as to promote the proper media circulation of business events.

With the prevalence of mobile devices such as smartphones and tablets, the ways people access to the Internet have changed enormously. In addition to the information that can be recorded by traditional Web-based e-commerce like frequent online shopping stores and browsing histories, mobile devices are capable of tracking sophisticated browsing behavior. The aim of this study is to utilize users' browsing behavior of reading hotel reviews on mobile devices and subsequently apply text-mining techniques to construct user interest profiles to make personalized hotel recommendations. Specifically, we design and implement an app where the user can search hotels and browse hotel reviews, and every gesture the user has performed on the touch screen when reading the hotel reviews is recorded. We then identify the paragraphs of hotel reviews that a user has shown interests based on the gestures the user has performed. Text mining techniques are applied to construct the interest profile of the user according to the review content the user has seriously read. We collect more than 5,000 reviews of hotels in Taipei, the largest metropolitan area of Taiwan, and recruit 18 users to participate in the experiment. Experimental results demonstrate that the recommendations made by our system better match the user's hotel selections than previous approaches.

The Support Vector Machines (SVMs) have been widely used for classification due to its ability to give low generalization error. In many practical applications of classification, however, the wrong prediction of a certain class is much severer than that of the other classes, making the original SVM unsatisfactory. In this paper, we propose the notion of Asymmetric Support Vector Machine (ASVM), an asymmetric extension of the SVM, for these applications. Different from the existing SVM extensions such as thresholding and parameter tuning, ASVM employs a new objective that models the imbalance between the costs of false predictions from different classes in a novel way such that user tolerance on false-positive rate can be explicitly specified. Such a new objective formulation allows us of obtaining a lower false-positive rate without much degradation of the prediction accuracy or increase in training time. Furthermore, we show that the generalization ability is preserved with the new objective. We also study the effects of the parameters in ASVM objective and address some implementation issues related to the Sequential Minimal Optimization (SMO) to cope with large-scale data. An extensive simulation is conducted and shows that ASVM is able to yield either noticeable improvement in performance or reduction in training time as compared to the previous arts.

Training support vector machines (SVMs) with nonlinear kernel functions on large-scale data are usually very time-consuming. In contrast, there exist faster solvers to train the linear SVM. We propose a technique which sufficiently approximates the infinite-dimensional implicit feature mapping of the Gaussian kernel function by a low-dimensional feature mapping. By explicitly mapping data to the low-dimensional features, efficient linear SVM solvers can be applied to train the Gaussian kernel SVM, which leverages the efficiency of linear SVM solvers to train a nonlinear SVM. Experimental results show that the proposed technique is very efficient and achieves comparable classification accuracy to a normal nonlinear SVM solver.

The support vector machine (SVM) is a widely used tool in classification problems. The SVM trains a classifier by solving an optimization problem to decide which instances of the training data set are support vectors, which are the necessarily informative instances to form the SVM classifier. Since support vectors are intact tuples taken from the training data set, releasing the SVM classifier for public use or shipping the SVM classifier to clients will disclose the private content of support vectors. This violates the privacy-preserving requirements for some legal or commercial reasons. The problem is that the classifier learned by the SVM inherently violates the privacy. This privacy violation problem will restrict the applicability of the SVM. To the best of our knowledge, there has not been work extending the notion of privacy preservation to tackle this inherent privacy violation problem of the SVM classifier. In this paper, we exploit this privacy violation problem, and propose an approach to postprocess the SVM classifier to transform it to a privacy-preserving classifier which does not disclose the private content of support vectors. The postprocessed SVM classifier without exposing the private content of training data is called Privacy-Preserving SVM Classifier (abbreviated as PPSVC). The PPSVC is designed for the commonly used Gaussian kernel function. It precisely approximates the decision function of the Gaussian kernel SVM classifier without exposing the sensitive attribute values possessed by support vectors. By applying the PPSVC, the SVM classifier is able to be publicly released while preserving privacy. We prove that the PPSVC is robust against adversarial attacks. The experiments on real data sets show that the classification accuracy of the PPSVC is comparable to the original SVM classifier.

Outsourcing the training of support vector machines (SVM) to external service providers benefits the data owner who is not familiar with the techniques of the SVM or has limited computing resources. In outsourcing, the data privacy is a critical issue for some legal or commercial reasons since there may be sensitive information contained in the data. Existing privacy-preserving SVM works are either not applicable to outsourcing or weak in security. In this paper, we propose a scheme for privacy-preserving outsourcing the training of the SVM without disclosing the actual content of the data to the service provider. In the proposed scheme, the data sent to the service provider is perturbed by a random transformation, and the service provider trains the SVM for the data owner from the perturbed data. The proposed scheme is stronger in security than existing techniques, and incurs very little redundant communication and computation cost.

Support vector machine (SVM) is a widely used tool in classification problem. SVM solves a quadratic optimization problem to decide which instances of training dataset are support vectors, i.e., the necessarily informative instances to form the classifier. The support vectors are intact tuples taken from the training dataset. Releasing the SVM classifier to public use or shipping the SVM classifier to clients will disclose the private content of support vectors, violating the privacy-preservation requirement in some legal or commercial reasons. To the best of our knowledge, there has not been work extending the notion of privacy-preservation to releasing the SVM classifier. In this paper, we propose an approximation approach which post-processes the SVM classifier to protect the private content of support vectors. This approach is designed for the commonly used Gaussian radial basis function kernel. By applying this post-processor on the SVM classifier, the resulted privacy-preserving SVM classifier can be publicly released without exposing the private content of support vectors and is able to provide comparable classification accuracy to the original SVM classifier.

Many practical applications of classification require the classifier to produce a very low false-positive rate. Although the Support Vector Machine (SVM) has been widely applied to these applications due to its superiority in handling high dimensional data, there are relatively little effort other than setting a threshold or changing the costs of slacks to ensure the low false-positive rate. In this paper, we propose the notion of Asymmetric Support VectorMachine (ASVM) that takes into account the false-positives and the user tolerance in its objective. Such a new objective formulation allows us to raise the confidence in predicting the positives, and therefore obtain a lower chance of false-positives. We study the effects of the parameters in ASVM objective and address some implementation issues related to the Sequential Minimal Optimization (SMO) to cope with large-scale data. An extensive simulation is conducted and shows that ASVM is able to yield either noticeable improvement in performance or reduction in training time as compared to the previous arts.

We explore in this paper a novel sampling algorithm, referred to as algorithm PAS (standing for proportion approximation sampling), to generate a high-quality online sample with the desired sample rate. The sampling quality refers to the consistency between the population proportion and the sample proportion of each categorical value in the database. Note that the state-of-the-art sampling algorithm to preserve the sampling quality has to examine the population proportion of each categorical value in a pilot sample a priori and is thus not applicable to incremental mining applications. To remedy this, algorithm PAS adaptively determines the inclusion probability of each incoming tuple in such a way that the sampling quality can be sequential/preserved while also guaranteeing the sample rate close to the user specified one. Importantly, PAS not only guarantees the proportion consistency of each categorical value but also excellently preserves the proportion consistency of multivariate statistics, which will be significantly beneficial to various data mining applications. For better execution efficiency, we further devise an algorithm, called algorithm EQAS (standing for efficient quality-aware sampling), which integrates PAS and random sampling to provide the flexibility of striking a compromise between the sampling quality and the sampling efficiency. As validated in experimental results on real and synthetic data, algorithm PAS can stably provide high-quality samples with corresponding computational overhead, whereas algorithm EQAS can flexibly generate samples with the desired balance between sampling quality and sampling efficiency