Search

Deep learning (DL) algorithms have played a major role in achieving state-of-the-art (SOTA) performance in various learning applications, including computer vision, natural language processing, and recommendation systems (RSs). However, these methods are based on a vast amount of data and do not perform as well when there is a limited amount of data available. Moreover, some of these applications (e.g., RSs) suffer from other issues such as data sparsity and the cold-start problem. While recent research on RSs used DL models based on side information (SI) (e.g., product reviews, film plots, etc.) to tackle these challenges, we propose boosting neural network (BNN), a new DL framework for capturing complex patterns, which requires just a limited amount of data. Unlike conventional boosting, BNN does not sum the predictions generated by its components. Instead, it uses these predictions as new SI features which enhances accuracy. Our framework can be utilized for many problems, including classification, regression, and ranking. In this paper, we demonstrate BNN's use for addressing a classification task. Comprehensive experiments conducted to illustrate BNN's effectiveness on three real-world datasets demonstrated its ability to outperform existing SOTA models for classification tasks (e.g., clickthrough rate prediction).

In this paper, we present a generic, query-efficient black-box attack against API call-based machine learning malware classifiers. We generate adversarial examples by modifying the malware's API call sequences and non-sequential features (printable strings), and these adversarial examples will be misclassified by the target malware classifier without affecting the malware's functionality. In contrast to previous studies, our attack minimizes the number of malware classifier queries required. In addition, in our attack, the attacker must only know the class predicted by the malware classifier; attacker knowledge of the malware classifier's confidence score is optional. We evaluate the attack effectiveness when attacks are performed against a variety of malware classifier architectures, including recurrent neural network (RNN) variants, deep neural networks, support vector machines, and gradient boosted decision trees. Our attack success rate is around 98% when the classifier's confidence score is known and 64% when just the classifier's predicted class is known. We implement four state-of-the-art query-efficient attacks and show that our attack requires fewer queries and less knowledge about the attacked model's architecture than other existing query-efficient attacks, making it practical for attacking cloud-based malware classifiers at a minimal cost., Accepted as a conference paper at ACSAC 2020

Automatic machine learning (AutoML) aims to automate the different aspects of the data science process and, by extension, allow non-experts to utilize "off the shelf" machine learning solution. One of the more popular AutoML methods is the Tree-based Pipeline Optimization Tool (TPOT), which uses genetic programming (GP) to efficiently explore the vast space of ML pipelines and produce a working ML solution. However, TPOT's GP process comes with substantial time and computational costs. In this study, we explore TPOT's GP process and propose MetaTPOT, an enhanced variant that uses a meta learning-based approach to predict the performance of TPOT's pipeline candidates. MetaTPOT leverages domain knowledge in the form of pipelines pre-ranking to improve TPOT's speed and performance. Evaluation on 65 classification datasets shows that our approach often improves the outcome of the genetic process while simultaneously substantially reduce its running time and computational cost.

The identification of criminals' behavioral patterns can be helpful for solving crimes. Currently, in order to perform this task, police investigators manually extract criminals' behavioral patterns (also referred to as criminals' modus operandi) from a large corpus of police reports. These patterns are compared to the patterns observed in an ongoing criminal investigation to identify similarities that may link the suspect to other documented crimes. Due to the large number of historical cases, this manual process is time consuming, very costly in terms of police resources, and limits the investigators' ability to solve open cases. In this study, we propose an automatic and language independent method for extracting behavioral patterns from police reports. Relying on the extracted behavioral patterns as input, we utilize a Siamese neural network to identify burglaries committed by the same criminals. Experiments performed using a large dataset of police reports written in Hebrew provided by the Israel Police demonstrate the proposed method's high performance, achieving an AUC above 0.9. Using our method, we are also able to identify potential suspects for 22.41% of the open burglary cases in Israel.

Temporal dynamic graphs are graphs whose topology evolves over time, with nodes and edges added and removed between different time snapshots. Embedding such graphs in a low-dimensional space is important for a variety of tasks, including graphs' similarities, time series trends analysis and anomaly detection, graph visualization, graph classification, and clustering. Despite the importance of the temporal element in these tasks, existing graph embedding methods focus on capturing the graph's nodes in a static mode and/or do not model the graph in its entirety in temporal dynamic mode. In this study, we present tdGraphEmbed, a novel temporal graph-level embedding approach that extend the random-walk based node embedding methods to globally embed both the nodes of the graph and its representation at each time step, thus creating representation of the entire graph at each step. Our approach was applied to graph similarity ranking, temporal anomaly detection, trend analysis, and graph visualizations tasks, where we leverage our temporal embedding in a fast and scalable way for each of the tasks. An evaluation of tdGraphEmbed on five real-world datasets shows that our approach can outperform state-of-the-art approaches used for graph embedding and node embedding in temporal graphs.

Many sellers on e-commerce platforms offer buyers product bundles, which package together two or more different items. The identification of such bundles is a necessary step to support a variety of related services, from recommendation to dynamic pricing. In this work, we present a comprehensive study of bundle identification on a large e-commerce website. Our analysis of bundle compared to non-bundle listed items reveals several key differentiating characteristics, spanning the listing's title, image, and attributes. Following, we experiment with a multi-modal classifier, which takes advantage of these characteristics as features. Our analysis also shows that a bundle indicator input by sellers tends to be highly noisy and carries only a weak signal. The bundle identification task therefore faces the challenge of having a small set of manually-labeled clean examples and a larger set of noisy-labeled examples, in conjunction with class imbalance due to the relative scarcity of bundles. Our experiments with basic supervised classifiers, using the manually-labeled and/or the noisy-labeled data for training, demonstrates only moderate performance. We therefore turn to a semisupervised approach and propose GREED, a self-training ensemblebased algorithm with a greedy model selection. Our evaluation over two different meta-categories shows a superior performance of semi-supervised approaches for the bundle identification task, with GREED outperforming several semi-supervised alternatives. The combination of textual, image, and some metadata features is shown to yield the best performance, reaching an AUC of 0.89 and 0.92 for the two meta-categories, respectively

One of the challenges of automating machine learning applications is the automatic selection of an algorithmic model for a given problem. We present AutoDi, a novel and resource-efficient approach for model selection. Our approach combines two sources of information: metafeatures extracted from the data itself and word-embedding features extracted from a large corpus of academic publications. This hybrid approach enables AutoDi to select top-performing algorithms both for widely and rarely used datasets by utilizing its two types of feature sets. We demonstrate the effectiveness of our proposed approach on a large dataset of 119 datasets and 179 classification algorithms grouped into 17 families. We show that AutoDi can reach an average of 98.8% of optimal accuracy and select the optimal classification algorithm in 49.5% of all cases.

Inferring socio-demographic attributes of users is an important and challenging task that could help with personalization, recommendation, advertising, etc. Sensor data collected from mobile devices can be utilized for inferring such attributes. Previous works have focused on combining different types of sensors, such as applications, accelerometer, GPS, battery, and many others, to achieve this task. In this study, we were able to infer attributes, such as gender, age, marital status, and whether the user has children, using solely the GPS sensor. We suggest a novel inference technique, which learns an embedding representation of preprocessed spatial GPS trajectories using an adaption of the Word2vec approach. Based on the embedding representation, we later train multiple classification models to achieve the inference goals. Our empirical results indicate that the suggested embedding approach outperforms a classification approach which does not take into consideration the embedding patterns. Experiments on real datasets collected from Android devices show that the proposed method achieves over 80% accuracy for various demographic prediction tasks.

Questions on community question answering websites usually reflect one of two intents: learning information or starting a conversation. In this paper, we revisit this fundamental classification task of informational versus conversational questions, which was originally introduced and studied in 2009. We use a substantially larger dataset of archived questions from Yahoo Answers, which includes the question»s title, description, answers, and votes. We replicate the original experiments over this dataset, point out the common and different from the original results, and present a broad set of characteristics that distinguish the two question types. We also develop new classifiers that make use of additional data types, advanced machine learning, and a large dataset of unlabeled data, which achieve enhanced performance.

Smartphones have become ubiquitous in our daily lives; they are used for a wide range of tasks and store increasing amounts of personal data. To minimize risk and prevent misuse of this data by unauthorized users, access must be restricted to verified users. Current classification-based methods for gesture-based user verification only consider single gestures, and not sequences. In this paper, we present a method which utilizes information from sequences of touchscreen gestures, and the context in which the gestures were made. To evaluate our approach, we built an application which records all the necessary data from the device (touch and contextual sensors which do not consume significant battery life), and installed it on several Galaxy S4 smartphones. The smartphones were given to 20 volunteers to use as their personal phones for two-weeks. Using XGBoost on the collected data, we were able to classify between a legitimate user and the population of illegitimate users (imposters) with an average equal error rate (EER) of 4.78% and an average area under the curve (AUC) of 98.15%. Our method demonstrates that by considering sequences of gestures, as opposed to individual gestures, the accuracy of the verification process improves significantly.

Context-aware systems enable the sensing and analysis of user context in order to provide personalized services. Our study is part of growing research efforts examining how high-dimensional data collected from mobile devices can be utilized to infer users' dynamic preferences. We present a novel method for inferring contextual user preferences by using an unsupervised deep learning technique applied to mobile sensor data. We train an auto-encoder for each user preference with contextual data that based on past user interaction with the system. Given new contextual sensor data from a user, the patterns discovered from each auto-encoder are used to predict the most likely preference in the given context. This can greatly enhance a variety of services, such as mobile online advertising and context-aware recommender systems. We demonstrate our contribution with a point of interest (POI) recommender system in which we label contextual preferences based on the interaction of users with categories of items. Empirical results utilizing a real world dataset of mobile users show a significant improvement (16% to 73% improvement) in classification accuracy compared with state of the art classification methods.

Recent methods for learning vector space representations of words, word embedding, such as GloVe and Word2Vec have succeeded in capturing fine-grained semantic and syntactic regularities. We analyzed the effectiveness of these methods for e-commerce recommender systems by transferring the sequence of items generated by users' browsing journey in an e-commerce website into a sentence of words. We examined the prediction of fine-grained item similarity (such as item most similar to iPhone 6 64GB smart phone) and item analogy (such as iPhone 5 is to iPhone 6 as Samsung S5 is to Samsung S6) using real life users' browsing history of an online European department store. Our results reveal that such methods outperform related models such as singular value decomposition (SVD) with respect to item similarity and analogy tasks across different product categories. Furthermore, these methods produce a highly condensed item vector space representation, item embedding, with behavioral meaning sub-structure. These vectors can be used as features in a variety of recommender system applications. In particular, we used these vectors as features in a neural network based models for anonymous user recommendation based on session's first few clicks. It is found that recurrent neural network that preserves the order of user's clicks outperforms standard neural network, item-to-item similarity and SVD (recall@10 value of 42% based on first three clicks) for this task.

In this paper we describe and share with the research community, a significant smartphone dataset obtained from an ongoing long-term data collection experiment. The dataset currently contains 10 billion data records from 30 users collected over a period of 1.6 years and an additional 20 users for 6 months (totaling 50 active users currently participating in the experiment).The experiment involves two smartphone agents: SherLock and Moriarty. SherLock collects a wide variety of software and sensor data at a high sample rate. Moriarty perpetrates various attacks on the user and logs its activities, thus providing labels for the SherLock dataset.The primary purpose of the dataset is to help security professionals and academic researchers in developing innovative methods of implicitly detecting malicious behavior in smartphones. Specifically, from data obtainable without superuser (root) privileges. To demonstrate possible uses of the dataset, we perform a basic malware analysis and evaluate a method of continuous user authentication.

Security systems for databases produce numerous alerts about anomalous activities and policy rule violations. Prioritizing these alerts will help security personnel focus their efforts on the most urgent alerts. Currently, this is done manually by security experts that rank the alerts or define static risk scoring rules. Existing solutions are expensive, consume valuable expert time, and do not dynamically adapt to changes in policy. Adopting a learning approach for ranking alerts is complex due to the efforts required by security experts to initially train such a model. The more features used, the more accurate the model is likely to be, but this will require the collection of a greater amount of user feedback and prolong the calibration process. In this paper, we propose CyberRank, a novel algorithm for automatic preference elicitation that is effective for situations with limited experts' time and outperforms other algorithms for initial training of the system. We generate synthetic examples and annotate them using a model produced by Analytic Hierarchical Processing (AHP) to bootstrap a preference learning algorithm. We evaluate different approaches with a new dataset of expert ranked pairs of database transactions, in terms of their risk to the organization. We evaluated using manual risk assessments of transaction pairs, CyberRank outperforms all other methods for cold start scenario with error reduction of 20%.

We believe that Deep Learning is one of the next big things in Recommendation Systems technology. The past few years have seen the tremendous success of deep neural networks in a number of complex tasks such as computer vision, natural language processing and speech recognition. Despite this, only little work has been published on Deep Learning methods for Recommender Systems. Notable recent application areas are music recommendation, news recommendation, and session-based recommendation. The aim of the workshop is to encourage the application of Deep Learning techniques in Recommender Systems, to promote research in deep learning methods for Recommender Systems, and to bring together researchers from the Recommender Systems and Deep Learning communities.

The query-performance prediction task aims at estimating the retrieval effectiveness of queries without obtaining relevance feedback from users. Most of the recently proposed predictors were empirically evaluated with various datasets to demonstrate their merits. We propose a framework for theoretical categorization and estimation of the value of query performance predictors (QPP) without empirical evaluation. We demonstrate the application of the proposed framework on four representative selected predictors and show how it emphasizes their strengths and weaknesses. The main contribution of this work is the theoretical grounded categorization of representative QPP.

Proactive recommender systems push recommendations to users without their explicit request whenever a recommendation that suits a user is available. These systems strive to optimize the match between recommended items and users' preferences. We assume that recommendations might be reflected with low accuracy not only due to the recommended items' suitability to the user, but also because of the recommendations' timings. We therefore claim that it is possible to learn a model of good and bad contexts for recommendations that can later be integrated in a recommender system. Using mobile data collected during a three week user study, we suggest a two-phase model that is able to classify whether a certain context is at all suitable for any recommendation, regardless of its content. Results reveal that a hybrid model that first decides whether it should use a personal or a non-personal timing model, and then classifies accordingly whether the timing is proper for recommendations, is superior to both the personal or non-personal timing models.

The 2015 ACM Recommender Systems Challenge offered the opportunity to work on a large-scale e-commerce dataset from a big retailer in Europe which is accepting recommender system as a service from YOOCHOOSE. Participants tackled the problem of predicting what items a user intends to purchase, if any, given a click sequence performed during an activity session on the e-commerce website. The challenge ran for seven months and was very successful, attracting 850 teams from 49 countries which submitted a total of 5,437 solutions. The winners of the challenge scored approximately 50% of the maximum score, which we considered as an impressive achievement. In this paper we provide a brief overview of the challenge and its results.

Online health communities are a major source for patients and their informal caregivers in the process of gathering information and seeking social support. The Cancer Survivors Network of the American Cancer Society has many users and presents a large number of user interactions with regards to coping with cancer. Sentiment analysis is an important process in understanding members' needs and concerns and the impact of users' responses on other members. It aims to determine the participants' subjective attitude and reflect their emotions. Analyzing the sentiment of posts in online health communities enables the investigation of various factors such as what affects the sentiment change and discovery of sentiment change patterns. Since each writer has his or her own personality, and temporal emotional state, behavioral traits can be reflected in the writer's writing style. Pronouns are function-words which often convey some unique styling patterns into the texts. Drawing on a lexical approach to emotions, we conduct factor analysis on the use of pronouns in self-descriptions texts. Our analysis shows that the usage of pronouns has an effect on sentiment classification. Moreover, we evaluated the use of pronouns in our domain, and found it different than standard English usage.

A group may appreciate recommendations on items that fit their joint preferences. When the members' actual preferences are unknown, a recommendation can be made with the aid of collaborative filtering methods. We offer to narrow down the recommended list of items by eliciting the users' actual preferences. Our final goal is to output top-k preferred items to the group out of the top-N recommendations provided by the recommender system (k

Context-aware computing aims at tailoring services to the user's circumstances and surroundings. Our study examines how data collected from mobile devices can be utilized to infer users' behavior and environment. We present the results and the lessons learned from a two-week user study of 40 students. The data collection was performed using Contexto, a framework for collecting data from a rich set of sensors installed on mobile devices, which was developed for this purpose. We studied various new and fine-grained user contexts which are relevant to students' daily activities, such as "in class and interested in the learned materials" and "on my way to campus". These contexts might later be utilized for various purposes such as recommending relevant items to the students' context. We compare various machine learning methods and report their effectiveness for the purposes of inferring the users' context from the collected data. In addition, we present our findings on how to evaluate context inference systems, on the importance of explicit and latent labeling for context inference and on the effect of new users on the results' accuracy.

The query-performance prediction task is to estimate retrieval effectiveness with no relevance judgments. Pre-retrieval prediction methods operate prior to retrieval time. Hence, these predictors are often based on analyzing the query and the corpus upon which retrieval is performed. We propose a {\em corpus-independent} approach to pre-retrieval prediction which relies on information extracted from Wikipedia. Specifically, we present Wikipedia-based features that can attest to the effectiveness of retrieval performed in response to a query {\em regardless} of the corpus upon which search is performed. Empirical evaluation demonstrates the merits of our approach. As a case in point, integrating the Wikipedia-based features with state-of-the-art pre-retrieval predictors that analyze the corpus yields prediction quality that is consistently better than that of using the latter alone.

Users of scientific papers databases, such as CiteSeer, Google Scholar, and Microsoft Academic, often search for papers using a set of keywords. Unfortunately, many authors avoid listing sufficient keywords for their papers. As such, these applications may need to automatically associate good descriptive keywords with papers. This is a well-studied problem given the complete text of the paper, but in many cases, due to copyright privileges, research papers databases do not have the complete text, only metadata, such as the title and abstract. On the other hand, research papers databases typically maintain the citation network of each paper. In this paper we study the problem of predicting which keywords are appropriate for a scientific paper, using only the citation network. We compare our method with predicting keywords using the title and abstract, concluding that the citation network provides much better predictions.

In this paper, we present a method for boosting collaborative filtering by integrating spatial information about geo-referenced items (e.g., photos). In particular, we developed a method to estimate missing ratings by propagating an item's neighbor's ratings based on the similarity of geospatial information. An empirical evaluation shows that geospatial information significantly improves recommendation results, and its contribution grows with the ratings data's level of sparseness. We illustrate the usefulness of the method for a photo recommendation task using data obtained from two popular photo-sharing websites: Flickr and Panoramio. A comparison with state-of-the-art methods indicates the superiority of the proposed method, implying that geospatial information should be considered, when available.

Academic papers contain multiple figures representing important findings and experimental results; we present a search engine specifically focused on figures in academic documents. This search engine allows users to search on figures in approximately 150,000 chemistry journal articles though the method is easily extendable to other domains. Our system indexes figure caption and mentions extracted from the PDF in documents using a custom built extractor. Recall and precision performance of extracted figures is in the 80 to 90% range. We give the frame work for the extraction algorithm, architecture and ranking function.

Insurance riders are optional addendum to base insurance policies. In this paper we discuss the application of recommender systems to the task of matching riders to clients. This task is difficult because of the variety of possible riders, as well as the poor knowledge of the client over these riders. We focus on call centers where the agent also has limited knowledge and expertise. For such agents, discovering appropriate riders for the current client is very difficult, and automated tools that suggest such riders can play an important role in the agent-client dialogue, and may influence considerably the outcome of the interaction.This paper presents and discusses in detail the problem of recommending insurance riders to clients in call centers, comparing it to other, classic, recommendation system applications. In addition, we present an analysis of customer purchase behavior, showing that simple item-item recommendation algorithms provide good recommendations for riders given a base policy.

Most collaborative Recommender Systems (RS) operate in a single domain (such as movies, books, etc.) and are capable of providing recommendations based on historical usage data which is collected in the specific domain only. Cross-domain recommenders address the sparsity problem by using Machine Learning (ML) techniques to transfer knowledge from a dense domain into a sparse target domain. In this paper we propose a transfer learning technique that extracts knowledge from multiple domains containing rich data (e.g., movies and music) and generates recommendations for a sparse target domain (e.g., games). Our method learns the relatedness between the different source domains and the target domain, without requiring overlapping users between domains. The model integrates the appropriate amount of knowledge from each domain in order to enrich the target domain data. Experiments with several datasets reveal that, using multiple sources and the relatedness between domains improves accuracy of results.

When we write or prepare to write a research paper, we always have appropriate references in mind. However, there are most likely references we have missed and should have been read and cited. As such a good citation recommendation system would not only improve our paper but, overall, the efficiency and quality of literature search. Usually, a citation's context contains explicit words explaining the citation. Using this, we propose a method that "translates" research papers into references. By considering the citations and their contexts from existing papers as parallel data written in two different "languages", we adopt the translation model to create a relationship between these two "vocabularies". Experiments on both CiteSeer and CiteULike dataset show that our approach outperforms other baseline methods and increase the precision, recall and f-measure by at least 5% to 10%, respectively. In addition, our approach runs much faster in the both training and recommending stage, which proves the effectiveness and the scalability of our work.

One important challenge in the field of recommender systems is the sparsity of available data. This problem limits the ability of recommender systems to provide accurate predictions of user ratings. We overcome this problem by using the publicly available user generated information contained in Wikipedia. We identify similarities between items by mapping them to Wikipedia pages and finding similarities in the text and commonalities in the links and categories of each page. These similarities can be used in the recommendation process and improve ranking predictions. We find that this method is most effective in cases where ratings are extremely sparse or nonexistent. Preliminary experimental results on the MovieLens dataset are encouraging.

In this demo we present a recommender benchmark framework that serves as an infrastructure for comparing and examining the performance and feasibility of different recommender algorithms on various datasets with a variety of measures. The extendable infrastructure aims to provide easy plugging of novel recommendation-algorithms, datasets and compare their performance using visual tools and metrics with other algorithms in the benchmark. It also aims at generating a WEKA-type workbench [1] for the recommender systems field to enable usage and application of common recommender systems (RS) algorithms for research and practice. The demo movie is available at: http://www.youtube.com/watch?v=fsDITf6s0WY

In this paper, we propose a new unsupervised approach for identifying suspicious access to sensitive relational data. In the proposed method, a tree-like model encapsulates the characteristics of the result-set (i.e., data) that the user normally access within each possible context. During the detection phase, result-sets are examined against the induced model and a similarity score is derived.

In previous work we proposed the M-score measure for assigning a misuseability (i.e., sensitivity) score to data records. The M-score uses sensitivity score functions that should be acquired from domain experts. In this paper we present two different approaches for acquiring the required knowledge. In the first method the expert is asked to explicitly assign a sensitivity score to displayed records. The second method employs pairwise comparison approach. A field study indicates that the later method is preferable.

Detecting data leakage/misuse poses a great challenge for organizations. Whether caused by malicious intent or an inadvertent mistake, data leakage/misuse can diminish a company's brand, reduce shareholder value, and damage the company's goodwill and reputation. This challenge is intensified when trying to detect and/or prevent data leakage/misuse performed by an insider with legitimate permissions to access the organization's systems and its critical data. In this paper we propose a new approach for identifying suspicious insiders who can access data stored in a database via an application. In the proposed method suspicious access to sensitive data is detected by analyzing the result-sets sent to the user following a request that the user submitted. Result-sets are analyzed within the instantaneous context in which the request was submitted. From the analysis of the result-set and the context we derive a "level of anomality". If the derived level is above a predefined threshold, an alert can be sent to the security officer. The proposed method applies data-linkage techniques in order to link the contextual features and the result-sets. Machine learning algorithms are then employed for generating a behavioral model during a learning phase. The behavioral model encapsulates knowledge on the behavior of a user; i.e., the characteristics of the result-sets of legitimate or malicious requests. This behavioral model is used for identifying malicious requests based on their abnormality. An evaluation with sanitized data shows the usefulness of the proposed method in detecting data misuse.

Over the past few years data leakage and data misuse have become a major concern for organizations. A data leakage or data misuse incident can damage an organization's reputation and brand name as well as compromise the privacy of its customers. Much research has been conducted in order to find a solution to these threats. Most methods are based on anomaly detection that tracks the user's behavior by examining the syntax of SQL queries in order to detect outlier queries. Other methods examine the data retrieved by the query. In this paper, we propose a new concept for analyzing the retrieved data - the Misuseability Weight. This approach focuses on assigning a score that represents the sensitivity level of the data exposed to the user. This measure predicts the ability of a user to exploit the exposed data in a malicious way. We suggest a new measure, the M-score, which assigns a misuseability weight to a table of data, propose some properties of the new measure and demonstrate its usefulness using over several leakage scenarios.

Group Recommendation Systems (GRS) aim at recommending items that are relevant for the joint interest of a group of users. Voting mechanisms assume that users rate all items in order to identify an item that suits the preferences of all group members. This assumption is not feasible in sparse rating scenarios which are common in the recommender systems domain. In this paper we examine an application of voting theory to GRS. We propose a method to accurately determine the winning item while using a minimal set of the group members ratings, assuming that the recommender system has probabilistic knowledge about the distribution of users' ratings of items in the system. Since computing the optimal minimal set of ratings is computationally intractable, we propose two heuristic algorithms that proceed iteratively that aiming atto minimizing the number of required ratings, until identifying a "winning item". Experiments with the Netflix data show that the proposed algorithms reduce the required number of ratings for identifying the "winning item" by more than 50%.

Recommender systems provide consumers with ratings of items. These ratings are based on a set of ratings that were obtained from a wide scope of users. Predicting the ratings can be formulated as a regression problem. Ensemble regression methods are effective tools that improve the results of simple regression algorithms by iteratively applying the simple algorithm to a diverse set of inputs. The present paper describes a simple and effective ensemble regressor for the prediction of missing ratings in recommender systems. The ensemble method is an adaptation of the AdaBoost regression algorithm for recommendation tasks. In all iterations, interpolation weights for all nearest neighbors are simultaneously derived by minimizing the root mean squared error. From iteration to iteration instances that are hard to predict are reinforced by manipulating their weights in the goal function that needs to be minimized. The experimental evaluation demonstrates that the ensemble methodology significantly improves the predictive performance of single neighborhood-based collaborative filtering.