Multi-armed bandit problems are receiving a great deal of attention because they adequately formalize the exploration-exploitation trade-offs arising in several industrially relevant applications, such as online advertisement and, more generally, recommendation systems. In many cases, however, these applications have a strong social component, whose integration in the bandit algorithm could lead to a dramatic performance increase. For instance, we may want to serve content to a group of users by taking advantage of an underlying network of social relationships among them. In this paper, we introduce novel algorithmic approaches to the solution of such networked bandit problems. More specifically, we design and analyze a global strategy which allocates a bandit algorithm to each network node (user) and allows it to “share” signals (contexts and payoffs) with the neghboring nodes. We then derive two more scalable variants of this strategy based on different ways of clustering the graph nodes. We experimentally compare the algorithm and its variants to state-of-the-art methods for contextual bandits that do not use the relational information. Our experiments, carried out on synthetic and real-world datasets, show a marked increase in prediction performance obtained by exploiting the network structure.

Automatic music recommendation has become an increasingly relevant problem in recent years, since a lot of music is now sold and consumed digitally. Most recommender systems rely on collaborative filtering. However, this approach suffers from the cold start problem: it fails when no usage data is available, so it is not effective for recommending new and unpopular songs. In this paper, we propose to use a latent factor model for recommendation, and predict the latent factors from music audio when they cannot be obtained from usage data. We compare a traditional approach using a bag-of-words representation of the audio signals with deep convolutional neural networks, and evaluate the predictions quantitatively and qualitatively on the Million Song Dataset. We show that using predicted latent factors produces sensible recommendations, despite the fact that there is a large semantic gap between the characteristics of a song that affect user preference and the corresponding audio signal. We also show that recent advances in deep learning translate very well to the music recommendation setting, with deep convolutional neural networks significantly outperforming the traditional approach.

One of the most challenging problems in recommender systems based on the collaborative filtering (CF) concept is data sparseness, i.e., limited user preference data is available for making recommendations. Cross-domain collaborative filtering (CDCF) has been studied as an effective mechanism to alleviate data sparseness of one domain using the knowledge about user preferences from other domains. A key question to be answered in the context of CDCF is what common characteristics can be deployed to link different domains for effective knowledge transfer. In this paper, we assess the usefulness of user-contributed (social) tags in this respect. We do so by means of the Generalized Tag-induced Cross-domain Collaborative Filtering (GTagCDCF) approach that we propose in this paper and that we developed based on the general collective matrix factorization framework. Assessment is done by a series of experiments, using publicly available CF datasets that represent three cross-domain cases, i.e., two two-domain cases and one three-domain case. A comparative analysis on two-domain cases involving GTagCDCF and several state-of-the-art CDCF approaches indicates the increased benefit of using social tags as representatives of explicit links between domains for CDCF as compared to the implicit links deployed by the existing CDCF methods. In addition, we show that users from different domains can already benefit from GTagCDCF if they only share a few common tags. Finally, we use the three-domain case to validate the robustness of GTagCDCF with respect to the scale of datasets and the varying number of domains.

This paper describes the Intelligent Engine (IE) of ECHOES, a serious game built for helping young children with Autism Spectrum Conditions acquire social communication skills. ECHOES IE's main component is an autonomous virtual agent that acts as a credible social partner for children with autism by engaging them in interactive learning activities. The other IE components are a user model, a drama manager and a social communication engine. We discuss how AI technology allows us to satisfy the requirements for the design of the agent and the learning activities that we identified through consultations with children and carers and a review of best practices for autism intervention. We present experimental results pertaining to the agent's effectiveness, which show encouraging improvements for a number of children.

With the growing use of automated content creation and computer-aided design tools in game development, there is potential for enhancing the design process through personalized interactions between the software and the game developer. This paper proposes designer modeling for capturing the designer's preferences, goals and processes from their interaction with a computer-aided design tool, and suggests methods and domains within game development where such a model can be applied. We describe how designer modeling could be integrated with current work on automated and mixed-initiative content creation, and envision future directions which focus on personalizing the processes to a designer's particular wishes.

Significant effort in machine learning and information retrieval has been devoted to identifying personalized content such as recommendations and search results. Personalized human computation has the potential to go beyond existing techniques like collaborative filtering to provide personal­ized results on demand, over personal data, and for complex tasks. This work-in-progress compares two approaches to personal­ized human computation. In both, users annotate a small set of training examples which are then used by the crowd to annotate unseen items. In the first approach, which we call taste-matching, crowd members are asked to annotate the same set of training examples, and the ratings of similar users on other items are then used to infer personal­ized ratings. In the second approach, taste-grokking, the crowd is presented with the training examples and asked to use them predict the ratings of the target user on other items.

Inferring user preferences over a set of items is an important problem that has found numerous applications. This work focuses on the scenario where the explicit feature representation of items is unavailable, a setup that is similar to collaborative filtering. In order to learn a user's preferences from his/her response to only a small number of pairwise comparisons, we propose to leverage the pairwise comparisons made by many crowd users, a problem we refer to as crowdranking. The proposed crowdranking framework is based on the theory of matrix completion, and we present efficient algorithms for solving the related optimization problem. Our theoretical analysis shows that, on average, only O ( r log m ) pairwise queries are needed to accurately recover the ranking list of m items for the target user, where r is the rank of the unknown rating matrix, r << m . Our empirical study with two real-world benchmark datasets for collaborative filtering and one crowdranking dataset we collected via Amazon Mechanical Turk shows the promising performance of the proposed algorithm compared to the state-of-the-art approaches.

This paper presents a case study of an online apparel retailer called Stitch Fix, which uses both machine and human processing in its recommendation system.

Multi-armed bandit problems are receiving a great deal of attention because they adequately formalize the exploration-exploitation trade-offs arising in several industrially relevant applications, such as online advertisement and, more generally, recommendation systems. In many cases, however, these applications have a strong social component, whose integration in the bandit algorithm could lead to a dramatic performance increase. For instance, we may want to serve content to a group of users by taking advantage of an underlying network of social relationships among them. In this paper, we introduce novel algorithmic approaches to the solution of such networked bandit problems. More specifically, we design and analyze a global strategy which allocates a bandit algorithm to each network node (user) and allows it to "share" signals (contexts and payoffs) with the neghboring nodes. We then derive two more scalable variants of this strategy based on different ways of clustering the graph nodes. We experimentally compare the algorithm and its variants to state-of-the-art methods for contextual bandits that do not use the relational information. Our experiments, carried out on synthetic and real-world datasets, show a marked increase in prediction performance obtained by exploiting the network structure.

If learning methods are to scale to the massive sizes of modern datasets, it is essential for the field of machine learning to embrace parallel and distributed computing. Inspired by the recent development of matrix factorization methods with rich theory but poor computational complexity and by the relative ease of mapping matrices onto distributed architectures, we introduce a scalable divide-and-conquer framework for noisy matrix factorization. We present a thorough theoretical analysis of this framework in which we characterize the statistical errors introduced by the "divide" step and control their magnitude in the "conquer" step, so that the overall algorithm enjoys high-probability estimation guarantees comparable to those of its base algorithm. We also present experiments in collaborative filtering and video background modeling that demonstrate the near-linear to superlinear speed-ups attainable with this approach.