Publications

Best Usage Context Prediction for Music Tracks

Recommender Systems (RSs) are software tools and techniques providing suggestions for items to be of use to a user. Often, better recommendations can be generated if the of the recommendation is known, e.g., in a music RS, the user mood or activity. However, to adapt the recommendations to the the dependency of the user preferences from the contextual conditions must be modeled. This requires explicit user evaluations/ratings for items in alternative contexts. In this work we investigate a novel approach for collecting and using contextually dependent ratings in recommender systems. We introduce the concept of \best context, i.e., the contextual conditions most suited for a particular item to be recommended. We designed an interface for collecting such data for music tracks. The collected data was then used to evaluate the quality of several \best context prediction methods based on user-to-user collaborative ltering. The results, in opposition to what we expected, show that the notion of best is user dependent. Moreover, among the approaches we tried, the best performing one uses a k-nearest neighbors classier where the user-to- user similarity measures the agreement of two users in assigning the best to items.

Improving Sentiment Analysis in an Online Cancer Survivor Community Using Dynamic Sentiment Lexicon

No abstract is provided for this article.

Machine Learning Prediction for Prognosis of Patients With Aortic Stenosis

Aortic valve stenosis of any degree is associated with poor outcomes.

Incorporating Fuzzy Logic in Data Mining Tasks

In this chapter we discuss how fuzzy logic extends the envelop of the main data mining tasks: clustering, classification, regression and association rules. We begin by presenting a formulation of the data mining using fuzzy logic attributes. Then, for each task, we provide a survey of the main algorithms and a detailed description (i.e. pseudo-code) of the most popular algorithms. There are two main types of uncertainty in supervised learning: statistical and cognitive. Statistical uncertainty deals with the random behavior of nature and all existing data mining techniques can handle the uncertainty that arises (or is assumed to arise) in the natural world from statistical variations or randomness. Cognitive uncertainty, on the other hand, deals with human cognition. Fuzzy set theory, first introduced by Zadeh in 1965, deals with cognitive uncertainty and seeks to overcome many of the problems found in classical set theory. For example, a major problem faced by researchers of control theory is that a small change in input results in a major change in output. This throws the whole control system into an unstable state. In addition there was also the problem that the representation of subjective knowledge was artificial and inaccurate. Fuzzy set theory is an attempt to confront these difficulties and in this chapter we show how it can be used in data mining tasks. Request access from your librarian to read this chapter's full text.

Automated algorithm selection using meta-learning and pre-trained deep convolution neural networks

Automated machine learning (AutoML), which explores the automation of various machine learning tasks, has seen significant progress in recent years. One of the fundamental tasks in this field is the selection of effective algorithms for a given dataset. This task is particularly challenging for tabular datasets where, unlike images or text, one cannot use pre-trained embeddings from other datasets. In this study, we present AutoIRAD, a novel meta-learning and vision-based approach for automatic classification algorithm selection for tabular datasets. Our approach is the first to generate image-based representations of entire tabular datasets, enabling us to model the features and interactions of all the dataset samples. Our image-based representation has two significant advantages. First, training our model across multiple datasets enables it to simultaneously learn from diverse domains and create more generalized and effective dataset representations. This representation can also be effectively applied to previously unseen datasets. Secondly, using images enables us to leverage large pre-trained networks (e.g., VGG). By tuning existing networks rather than training new ones, we can both speed up the training of our model and reduce the needed amounts of training data. An extensive evaluation of 150 datasets shows that AutoIRAD achieves results comparable to the leading state-of-the-art AutoML solution but at a fraction of their computation time.

Taxonomy of mobile users' security awareness

The popularity of smartphones, coupled with the amount of valuable and private information they hold, make them attractive to attackers interested in exploiting the devices to harvest sensitive information. Exploiting human vulnerabilities (i.e., social engineering) is an approach widely used to achieve this goal. Improving the security awareness of users is an effective method for mitigating social engineering attacks. However, while in the domain of personal computers (PCs) the security awareness of users is relatively high, previous studies have shown that for the mobile platform, the security awareness level is significantly lower. The skills required from a mobile user to interact safely with his/her smartphone are different from those that are required for safe and responsible PC use. Therefore, the awareness of mobile users to security risks is an important aspect of information security. An essential and challenging requirement of assessing security awareness is the definition of measureable criteria for a security aware user. In this paper, we present a hierarchical taxonomy for security awareness, specifically designed for mobile device users. The taxonomy defines a set of measurable criteria that are categorized according to different technological focus areas (e.g., applications and browsing) and within the context of psychological dimensions (e.g., knowledge, attitude, and behavior). We demonstrate the applicability of the proposed taxonomy by introducing an expert-based procedure for deriving mobile security awareness models for different attack classes (each class is an aggregation of social engineering attacks that exploit a similar set of human vulnerabilities). Each model reflects the contribution (weight) of each criterion to the mitigation of the corresponding attack class. Application of the proposed procedure, based on the input of 17 security experts, to derive mobile security awareness models of four different attack classes, confirms that the skills required from a smartphone user to mitigate an attack are different for different attack classes.

Boosting Simple Collaborative Filtering Models Using Ensemble Methods

In this paper we examine the effect of applying ensemble learning to the performance of collaborative filtering methods. We present several systematic approaches for generating an ensemble of collaborative filtering models based on a single collaborative filtering algorithm (single-model or homogeneous ensemble). We present an adaptation of several popular ensemble techniques in machine learning for the collaborative filtering domain, including bagging, boosting, fusion and randomness injection. We evaluate the proposed approach on several types of collaborative filtering base models: k- NN, matrix factorization and a neighborhood matrix factorization model. Empirical evaluation shows a prediction improvement compared to all base CF algorithms. In particular, we show that the performance of an ensemble of simple (weak) CF models such as k-NN is competitive compared with a single strong CF model (such as matrix factorization) while requiring an order of magnitude less computational cost.

Feature Selection by Combining Multiple Methods

No abstract is provided for this article.

Diffusion Ensemble Classifiers

Transfer Learning for Content-Based Recommender Systems using Tree Matching

In this paper we present a new approach to content-based transfer learning for solving the data sparsity problem in cases when the users' preferences in the target domain are either scarce or unavailable, but the necessary information on the preferences exists in another domain. We show that training a system to use such information across domains can produce better performance. Specifically, we represent users' behavior patterns based on topological graph structures. Each behavior pattern represents the behavior of a set of users, when the users' behavior is defined as the items they rated and the items' rating values. In the next step we find a correlation between behavior patterns in the source domain and behavior patterns in the target domain. This mapping is considered a bridge between the two domains. Based on the correlation and content-attributes of the items, we train a machine learning model to predict users' ratings in the target domain. When we compare our approach to the popularity approach and KNN-cross-domain on a real world dataset, the results show that on an average of 83$%$ of the cases our approach outperforms both methods.

Data Leakage/Misuse Scenarios

No abstract is provided for this article.

Identity theft, computers and behavioral biometrics

The increase of online services, such as eBanks, WebMails, in which users are verified by a username and password, is increasingly exploited by Identity Theft procedures. Identity Theft is a fraud, in which someone pretends to be someone else is order to steal money or get other benefits. To overcome the problem of Identity Theft an additional security layer is required. Within the last decades the option of verifying users based on their keystroke dynamics was proposed during login verification. Thus, the imposter has to be able to type in a similar way to the real user in addition to having the username and password. However, verifying users upon login is not enough, since a logged station/mobile is vulnerable for imposters when the user leaves her machine. Thus, verifying users continuously based on their activities is required. Within the last decade there is a growing interest and use of biometrics tools, however, these are often costly and require additional hardware. Behavioral biometrics, in which users are verified, based on their keyboard and mouse activities, present potentially a good solution. In this paper we discuss the problem of Identity Theft and propose behavioral biometrics as a solution. We survey existing studies and list the challenges and propose solutions.

A GIS-based decision support system for hotel room rate estimation and temporal price prediction: The hotel brokers' context

The vastly increasing number of online hotel room bookings is not only intensifying the competition in the travel industry as a whole, but also prompts travel intermediates (i.e. e-companies that aggregate information about different travel products from different travel suppliers) into a fierce competition for the best prices of travel products, i.e. hotel rooms. An important factor that affects revenues is the ability to conclude profitable deals with different travel suppliers. However, the profitability of a contract not only depends on the communication skills of a contract manager. It significantly depends on the objective information obtained about a specific travel supplier and his/her products. While the contract manager usually has a broad knowledge of the travel business in general, collecting and processing specific information about travel suppliers is usually a time and cost expensive task. Our goal is to develop a tool that assists the travel intermediate to acquire the missing strategic information about individual hotels in order to leverage profitable deals. We present a GIS-based decision support system that can both, estimate objective hotel room rates using essential hotel and locational characteristics and predict temporal room rate prices. Information about objective hotel room rates allows for an objective comparison and provides the basis for a realistic computation of the contract's profitability. The temporal prediction of room rates can be used for monitoring past hotel room rates and for adjusting the price of the future contract. This paper makes three major contributions. First, we present a GIS-based decision support system, the first of its kind, for hotel brokers. Second, the DSS can be applied to virtually any part of the world, which makes it a very attractive business tool in real-life situations. Third, it integrates a widely used data mining framework that provides access to dozens of ready to run algorithms to be used by a domain expert and it offers the possibility of adding new algorithms once they are developed. The system has been designed and evaluated in close cooperation with a company that develops travel technology solutions, in particular inventory management and pricing solutions for many well-known websites and travel agencies around the world. This company has also provided us with real, large datasets to evaluate the system. We demonstrate the functionality of the DSS using the hotel data in the area of Barcelona, Spain. The results indicate the potential usefulness of the proposed system.

Machine-learning analysis of factors that shape cancer aneuploidy landscapes reveals an important role for negative selection

Aneuploidy, an abnormal number of chromosomes within a cell, is considered a hallmark of cancer. Patterns of aneuploidy differ across cancers, yet are similar in cancers affecting closely-related tissues. The selection pressures underlying aneuploidy patterns are not fully understood, hindering our understanding of cancer development and progression. Here, we applied interpretable machine learning (ML) methods to study tissue-selective aneuploidy patterns. We defined 20 types of features of normal and cancer tissues, and used them to model gains and losses of chromosome-arms in 24 cancer types. In order to reveal the factors that shape the tissue-specific cancer aneuploidy landscapes, we interpreted the ML models by estimating the relative contribution of each feature to the models. While confirming known drivers of positive selection, our quantitative analysis highlighted the importance of negative selection for shaping the aneuploidy landscapes of human cancer. Tumor-suppressor gene density was a better predictor of gain patterns than oncogene density, and vice-versa for loss patterns. We identified the contribution of tissue-selective features and demonstrated them experimentally for chr13q gain in colon cancer. In line with an important role for negative selection in shaping the aneuploidy landscapes, we found compensation by paralogs to be a top predictor of chromosome-arm loss prevalence, and demonstrated this relationship for one such paralog interaction. Similar factors were found to shape aneuploidy patterns in human cancer cell lines, demonstrating their relevance for aneuploidy research. Overall, our quantitative, interpretable ML models improve the understanding of the genomic properties that shape cancer aneuploidy landscapes.

An evolutionary algorithm for constructing a decision forest: Combining the classification of disjoints decision trees

Decision forest is an ensemble classification method that combines multiple decision trees to in a manner that results in more accurate classifications. By combining multiple heterogeneous decision trees, decision forest is effective in mitigating noise that is often prevalent in real-world classification tasks. This paper presents a new genetic algorithm for constructing a decision forest. Each decision tree classifier is trained using a disjoint set of attributes. Moreover, we examine the effectiveness of using a Vapnik–Chervonenkis dimension bound for evaluating the fitness function of decision forest. The new algorithm was tested on various datasets. The obtained results have been compared to other methods, indicating the superiority of the proposed algorithm. © 2008 Wiley Periodicals, Inc.

Cost-Sensitive Detection of Malicious Applications in Mobile Devices

No abstract is provided for this article.

Evolving context-aware recommender systems with users in mind

A context-aware recommender system (CARS) utilizes users’ context to provide personalized services. Contextual information can be derived from sensors in order to improve the accuracy of the recommendations. In this work, we focus on CARSs with high-dimensional contextual information that typically impacts the recommendation model, for example, by increasing the model’s dimensionality and sparsity. Generating accurate recommendations is not enough to constitute a useful system from the user’s perspective, since the use of some contextual information may cause problems, such as draining the user’s battery, raising privacy concerns, and more. Previous studies suggested reducing the amount of contextual information utilized by using domain knowledge to select the most suitable information. This approach is only applicable when the set of contexts is small enough to handle and sufficient for preventing sparsity. Moreover, hand-crafted context information may not represent an optimal set of features for the recommendation process. Another approach is to compress the contextual information into a denser latent space, but this may limit the ability to explain the recommended items to the users or compromise their trust. In this paper, we present a multi-step approach for selecting low-dimensional subsets of contextual information and incorporating them explicitly within CARSs. At the core of our approach is a novel feature selection algorithm based on genetic algorithms, which outperforms state-of-the-art dimensionality reduction CARS algorithms by improving recommendation accuracy and interpretability. Over the course of evolution, thousands of diverse feature subsets are generated; a deep context-aware model is produced for each feature subset, and the subsets are stacked together. The resulting stacked model is accurate and only uses interpretable, explicit features. Our approach includes a mechanism of tuning the different underlying algorithms that affect user concerns, such as privacy and battery consumption. We evaluated our approach on two high-dimensional context-aware datasets derived from smartphones. An empirical analysis of our results confirms that our proposed approach outperforms state-of-the-art CARS models while improving transparency and interpretability for the user. In addition to the empirical results, we present several use cases, examples and methodology of how researchers, domain experts and CARS modelers can tweak the feature selection algorithm to improve various user concerns and interpretability.

Explainable decision forest: Transforming a decision forest into an interpretable tree

Decision forests are considered the best practice in many machine learning challenges, mainly due to their superior predictive performance. However, simple models like decision trees may be preferred over decision forests in cases in which the generated predictions must be efficient or interpretable (e.g. in insurance or health-related use cases). This paper presents a novel method for transforming a decision forest into an interpretable decision tree, which aims at preserving the predictive performance of decision forests while enabling efficient classifications that can be understood by humans. This is done by creating a set of rule conjunctions that represent the original decision forest; the conjunctions are then hierarchically organized to form a new decision tree. We evaluate the proposed method on 33 UCI datasets and show that the resulting model usually approximates the ROC AUC gained by random forest while providing an interpretable decision path for each classification.