Alexandros Karatzoglou

Multiverse recommendation: n-dimensional tensor factorization for context-aware collaborative filtering

Context has been recognized as an important factor to consider in personalized Recommender Systems. However, most model-based Collaborative Filtering approaches such as Matrix Factorization do not provide a straightforward way of integrating context information into the model. In this work, we introduce a Collaborative Filtering method based on Tensor Factorization, a generalization of Matrix Factorization that allows for a flexible and generic integration of contextual information by modeling the data as a User-Item-Context N-dimensional tensor instead of the traditional 2D User-Item matrix. In the proposed model, called Multiverse Recommendation, different types of context are considered as additional dimensions in the representation of the data as a tensor. The factorization of this tensor leads to a compact model of the data which can be used to provide context-aware recommendations. We provide an algorithm to address the N-dimensional factorization, and show that the Multiverse Recommendation improves upon non-contextual Matrix Factorization up to 30% in terms of the Mean Absolute Error (MAE). We also compare to two state-of-the-art context-aware methods and show that Tensor Factorization consistently outperforms them both in semi-synthetic and real-world data - improvements range from 2.5% to more than 12% depending on the data. Noticeably, our approach outperforms other methods by a wider margin whenever more contextual information is available.

CLiMF: learning to maximize reciprocal rank with collaborative less-is-more filtering

In this paper we tackle the problem of recommendation in the scenarios with binary relevance data, when only a few (k) items are recommended to individual users. Past work on Collaborative Filtering (CF) has either not addressed the ranking problem for binary relevance datasets, or not specifically focused on improving top-k recommendations. To solve the problem we propose a new CF approach, Collaborative Less-is-More Filtering (CLiMF). In CLiMF the model parameters are learned by directly maximizing the Mean Reciprocal Rank (MRR), which is a well-known information retrieval metric for measuring the performance of top-k recommendations. We achieve linear computational complexity by introducing a lower bound of the smoothed reciprocal rank metric. Experiments on two social network datasets demonstrate the effectiveness and the scalability of CLiMF, and show that CLiMF significantly outperforms a naive baseline and two state-of-the-art CF methods.

TFMAP: optimizing MAP for top-n context-aware recommendation

In this paper, we tackle the problem of top-N context-aware recommendation for implicit feedback scenarios. We frame this challenge as a ranking problem in collaborative filtering (CF). Much of the past work on CF has not focused on evaluation metrics that lead to good top-N recommendation lists in designing recommendation models. In addition, previous work on context-aware recommendation has mainly focused on explicit feedback data, i.e., ratings. We propose TFMAP, a model that directly maximizes Mean Average Precision with the aim of creating an optimally ranked list of items for individual users under a given context. TFMAP uses tensor factorization to model implicit feedback data (e.g., purchases, clicks) with contextual information. The optimization of MAP in a large data collection is computationally too complex to be tractable in practice. To address this computational bottleneck, we present a fast learning algorithm that exploits several intrinsic properties of average precision to improve the learning efficiency of TFMAP, and to ensure its scalability. We experimentally verify the effectiveness of the proposed fast learning algorithm, and demonstrate that TFMAP significantly outperforms state-of-the-art recommendation approaches.

Improving maximum margin matrix factorization

Maximum Margin Matrix Factorization (MMMF) has been proposed as a learning approach to the task of collaborative filtering with promising results. In our recent paper [2], we proposed to extend the general MMMF framework to allow for structured (ranking) losses in addition to the squared error loss. In this paper, we introduce a novel algorithm to compute the ordinal regression ranking loss which is significantly faster than the state of the art. In addition, we propose severals extensions to the MMMF model: We introduce offset terms to cater for user and item biases. Users exhibit vastly different rating frequencies ranging from only one rating per user to thousands of them. Similarly, some items get thousands of ratings while others get rated only once. We introduce an adaptive regularizer to allow for more complex models for those items and users with many ratings. Finally, we show equivalence between a recent extension introduced in and a graph kernel approach described in [3]. Both aim at providing meaningful predictions for users with very little training data by virtue of the recommender graph. We performed an evaluation of these extensions on two standard data sets: Eachmovie and Movielens. These experiments show that the introduced extensions do improve the predictive performance over the original MMMF formulation, even though we did not formally optimize the parameters.

Climbing the app wall: enabling mobile app discovery through context-aware recommendations

The explosive growth of the mobile application (app) market has made it difficult for users to find the most interesting and relevant apps from the hundreds of thousands that exist today. Context is key in the mobile space and so too are proactive services that ease user input and facilitate effective interaction. We believe that to enable truly novel mobile app recommendation and discovery, we need to support real context-aware recommendation that utilizes the diverse range of implicit mobile data available in a fast and scalable manner. In this paper we introduce the Djinn model, a novel context-aware collaborative filtering algorithm for implicit feedback data that is based on tensor factorization. We evaluate our approach using a dataset from an Android mobile app recommendation service called appazaar. Our results show that our approach compares favorably with state-of-the-art collaborative filtering methods.

Parallel Recurrent Neural Network Architectures for Feature-rich Session-based Recommendations

Real-life recommender systems often face the daunting task of providing recommendations based only on the clicks of a user session. Methods that rely on user profiles -- such as matrix factorization -- perform very poorly in this setting, thus item-to-item recommendations are used most of the time. However the items typically have rich feature representations such as pictures and text descriptions that can be used to model the sessions. Here we investigate how these features can be exploited in Recurrent Neural Network based session models using deep learning. We show that obvious approaches do not leverage these data sources. We thus introduce a number of parallel RNN (p-RNN) architectures to model sessions based on the clicks and the features (images and text) of the clicked items. We also propose alternative training strategies for p-RNNs that suit them better than standard training. We show that p-RNN architectures with proper training have significant performance improvements over feature-less session models while all session-based models outperform the item-to-item type baseline.

Gaussian process factorization machines for context-aware recommendations

Context-aware recommendation (CAR) can lead to significant improvements in the relevance of the recommended items by modeling the nuanced ways in which context influences preferences. The dominant approach in context-aware recommendation has been the multidimensional latent factors approach in which users, items, and context variables are represented as latent features in low-dimensional space. An interaction between a user, item, and a context variable is typically modeled as some linear combination of their latent features. However, given the many possible types of interactions between user, items and contextual variables, it may seem unrealistic to restrict the interactions among them to linearity. To address this limitation, we develop a novel and powerful non-linear probabilistic algorithm for context-aware recommendation using Gaussian processes. The method which we call Gaussian Process Factorization Machines (GPFM) is applicable to both the explicit feedback setting (e.g. numerical ratings as in the Netflix dataset) and the implicit feedback setting (i.e. purchases, clicks). We derive stochastic gradient descent optimization to allow scalability of the model. We test GPFM on five different benchmark contextual datasets. Experimental results demonstrate that GPFM outperforms state-of-the-art context-aware recommendation methods.

Coverage, redundancy and size-awareness in genre diversity for recommender systems

There is increasing awareness in the Recommender Systems field that diversity is a key property that enhances the usefulness of recommendations. Genre information can serve as a means to measure and enhance the diversity of recommendations and is readily available in domains such as movies, music or books. In this work we propose a new Binomial framework for defining genre diversity in recommender systems that takes into account three key properties: genre coverage, genre redundancy and recommendation list size-awareness. We show that methods previously proposed for measuring and enhancing recommendation diversity - including those adapted from search result diversification - fail to address adequately these three properties. We also propose an efficient greedy optimization technique to optimize Binomial diversity. Experiments with the Netflix dataset show the properties of our framework and comparison with state of the art methods.

Learning to rank for recommender systems

Recommender system aim at providing a personalized list of items ranked according to the preferences of the user, as such ranking methods are at the core of many recommendation algorithms. The topic of this tutorial focuses on the cutting-edge algorithmic development in the area of recommender systems. This tutorial will provide an in depth picture of the progress of ranking models in the field, summarizing the strengths and weaknesses of existing methods, and discussing open issues that could be promising for future research in the community. A qualitative and quantitative comparison between different models will be provided while we will also highlight recent developments in the areas of Reinforcement Learning.

xCLiMF: optimizing expected reciprocal rank for data with multiple levels of relevance

Extended Collaborative Less-is-More Filtering xCLiMF is a learning to rank model for collaborative filtering that is specifically designed for use with data where information on the level of relevance of the recommendations exists, e.g. through ratings. xCLiMF can be seen as a generalization of the Collaborative Less-is-More Filtering (CLiMF) method that was proposed for top-N recommendations using binary relevance (implicit feedback) data. The key contribution of the xCLiMF algorithm is that it builds a recommendation model by optimizing Expected Reciprocal Rank, an evaluation metric that generalizes reciprocal rank in order to incorporate user feedback with multiple levels of relevance. Experimental results on real-world datasets show the effectiveness of xCLiMF, and also demonstrate its advantage over CLiMF when more than two levels of relevance exist in the data.