Harald Steck

A survey of collaborative filtering based social recommender systems

Recommendation plays an increasingly important role in our daily lives. Recommender systems automatically suggest to a user items that might be of interest to her. Recent studies demonstrate that information from social networks can be exploited to improve accuracy of recommendations. In this paper, we present a survey of collaborative filtering (CF) based social recommender systems. We provide a brief overview over the task of recommender systems and traditional approaches that do not use social network information. We then present how social network information can be adopted by recommender systems as additional input for improved accuracy. We classify CF-based social recommender systems into two categories: matrix factorization based social recommendation approaches and neighborhood based social recommendation approaches. For each category, we survey and compare several representative algorithms.

Training and testing of recommender systems on data missing not at random

Users typically rate only a small fraction of all available items. We show that the absence of ratings carries useful information for improving the top-k hit rate concerning all items, a natural accuracy measure for recommendations. As to test recommender systems, we present two performance measures that can be estimated, under mild assumptions, without bias from data even when ratings are missing not at random (MNAR). As to achieve optimal test results, we present appropriate surrogate objective functions for efficient training on MNAR data. Their main property is to account for all ratings - whether observed or missing in the data. Concerning the top-k hit rate on test data, our experiments indicate dramatic improvements over even sophisticated methods that are optimized on observed ratings only.

KL-divergence kernel regression for non-Gaussian fingerprint based localization

Various methods have been developed for indoor localization using WLAN signals. Algorithms that fingerprint the Received Signal Strength Indication (RSSI) of WiFi for different locations can achieve tracking accuracies of the order of a few meters. RSSI fingerprinting suffers though from two main limitations: first, as the signal environment changes, so does the fingerprint database, which requires regular updates; second, it has been reported that, in practice, certain devices record more complex (e.g bimodal) distributions of WiFi signals, precluding algorithms based on the mean RSSI. In this article, we propose a simple methodology that takes into account the full distribution for computing similarities among fingerprints using Kullback-Leibler divergence, and that performs localization through kernel regression. Our method provides a natural way of smoothing over time and trajectories. Moreover, we propose unsupervised KL-divergence-based recalibration of the training fingerprints. Finally, we apply our method to work with histograms of WiFi connections to access points, ignoring RSSI distributions, and thus removing the need for recalibration. We demonstrate that our results outperform nearest neighbors or Kalman and Particle Filters, achieving up to 1m accuracy in office environments. We also show that our method generalizes to non-Gaussian RSSI distributions.

Probability kernel regression for WiFi localisation

Various methods have been developed for indoor localisation using WLAN signals. Algorithms that fingerprint the received signal strength indicators (RSSI) of WiFi for different locations can achieve tracking accuracies of the order of a few metres. RSSI fingerprinting suffers from two main limitations: first, as the signal environment changes, so does the fingerprint database, which requires regular updates; second, it has been reported that, in practice, certain devices record more complex (e.g bimodal) distributions of WiFi signals, precluding algorithms based on the mean RSSI. Mirowski et al. [2011. KL-divergence kernel regression for non-Gaussian fingerprint based localization. In: International conference on indoor positioning and indoor navigation, Guimaraes, Portugal] have recently introduced a simple methodology that takes into account the full distribution for computing similarities among fingerprints using the Kullback–Leibler (KL) divergence, and then performs localisation through kernel regression. Their algorithm provides a natural way of smoothing over time and motion trajectories and can be applied directly to histograms of WiFi connections to access points, ignoring RSSI distributions, hence removing the need for fingerprint recalibration. It has been shown to outperform nearest neighbours or Kalman and particle filtres, achieving up to 1 m accuracy in office environments. In this article, we focus on the relevance of Gaussian or non-Gaussian distributions for modelling RSSI distributions by considering additional probabilistic kernels for comparing Gaussian distributions and by evaluating them on three contrasting datasets. We discuss their limitations and formulate how the KL-divergence kernel regression algorithm bridges the gap with other WiFi localisation algorithms, notably Bayesian networks, support vector machines and K nearest neighbours. Finally, we revisit the assumptions on the fingerprint maps and overview practical WiFi localisation software implementation.

Gaussian Ranking by Matrix Factorization

The ranking quality at the top of the list is crucial in many real-world applications of recommender systems. In this paper, we present a novel framework that allows for pointwise as well as listwise training with respect to various ranking metrics. This is based on a training objective function where we assume that, for given a user, the recommender system predicts scores for all items that follow approximately a Gaussian distribution. We motivate this assumption from the properties of implicit feedback data. As a model, we use matrix factorization and extend it by non-linear activation functions, as customary in the literature of artificial neural networks. In particular, we use non-linear activation functions derived from our Gaussian assumption. Our preliminary experimental results show that this approach is competitive with state-of-the-art methods with respect to optimizing the Area under the ROC curve, while it is particularly effective in optimizing the head of the ranked list.

Interactive Recommender Systems: Tutorial

In this tutorial we will explore the field of interactive video and music recommendations and their application at Netflix and Spotify. Interactive recommender systems enable the user to steer the received recommendations in the desired direction through explicit interaction with the system. In this tutorial, we outline the various aspects that are crucial for a smooth and effective user experience. In particular, we present our insights from several A/B tests. The tutorial will help researchers and practitioners in the RecSys community to gain a deeper understanding of the challenges related to the application of recommender systems in the online video and music entertainment business.

Public space behavior modeling with video and sensor analytics

We present a review of technologies relevant to public space surveillance and describe a pilot study to explore the challenges. The general purpose of this study is to capture and analyze behavior patterns and anomalies of people behavior in a public space. On the capture side, we explore a small array of networked cameras as well as an ultrasonic sensor array for measuring the height of walking persons. After capture, video and ultrasound signals are analyzed and statistics calculated for such measurements, including the duration and speed of the trajectory of each tracked person, and a persons height which is a useful biometric feature for tracking the person across multiple, non-overlapping camera views. These statistics are first analyzed offline to determine the expected patterns of measured values over many captured events. Based on the expected patterns, anomalies can be detected as outliers in real time. Since this is a broad-based pilot study, conclusions relate to the effectiveness of the capture modalities and approaches investigated. We discuss how we use these findings to guide our future work.

REDD 2014 -- international workshop on recommender systems evaluation: dimensions and design

Evaluation is a cardinal issue in recommender systems; as in any technical discipline, it highlights to a large extent the problems that need to be solved by the field and, hence, leads the way for algorithmic research and development in the community. Yet, in the field of recommender systems, there still exists considerable disparity in evaluation methods, metrics and experimental designs, as well as a significant mismatch between evaluation methods in the lab and what constitutes an effective recommendation for real users and businesses. Even after the relevant quality dimensions have been defined, a clear evaluation protocol should be specified in detail and agreed upon, allowing for the comparison of results and experiments conducted by different authors. This would enable any contribution to the same problem to be incremental and add up on top of previous work, rather than grow sideways. The REDD 2014 workshop seeks to provide an informal forum to tackle such issues and to move towards better understood and shared evaluation methodologies, allowing one to leverage the efforts and the workforce of the academic community towards meaningful and relevant directions in real-world developments.

Calibrated recommendations

When a user has watched, say, 70 romance movies and 30 action movies, then it is reasonable to expect the personalized list of recommended movies to be comprised of about 70% romance and 30% action movies as well. This important property is known as calibration, and recently received renewed attention in the context of fairness in machine learning. In the recommended list of items, calibration ensures that the various (past) areas of interest of a user are reflected with their corresponding proportions. Calibration is especially important in light of the fact that recommender systems optimized toward accuracy (e.g., ranking metrics) in the usual offline-setting can easily lead to recommendations where the lesser interests of a user get crowded out by the users main interests-which we show empirically as well as in thought-experiments. This can be prevented by calibrated recommendations. To this end, we outline metrics for quantifying the degree of calibration, as well as a simple yet effective re-ranking algorithm for post-processing the output of recommender systems.