Yehuda Koren’s research while affiliated with Google Inc. and other places

Yahoo! Answers is currently one of the most popular question answering systems. We claim however that its user experience could be significantly improved if it could route the "right question" to the "right user." Indeed, while some users would rush answering a question such as "what should I wear at the prom?," others would be upset simply being exposed to it. We argue here that Community Question Answering sites in general and Yahoo! Answers in particular, need a mechanism that would expose users to questions they can relate to and possibly answer. We propose here to address this need via a multi-channel recommender system technology for associating questions with potential answerers on Yahoo! Answers. One novel aspect of our approach is exploiting a wide variety of content and social signals users regularly provide to the system and organizing them into channels. Content signals relate mostly to the text and categories of questions and associated answers, while social signals capture the various user interactions with questions, such as asking, answering, voting, etc. We fuse and generalize known recommendation approaches within a single symmetric framework, which incorporates and properly balances multiple types of signals according to channels. Tested on a large scale dataset, our model exhibits good performance, clearly outperforming standard baselines.

Most email applications devote a significant part of their real estate to organization mechanisms such as folders. Yet, we verified on the Yahoo! Mail service that 70% of email users have never defined a single folder. This implies that one of the most well known email features is underexploited. We propose here to revive the feature by providing a method for generating a lighter form of folders, or tags, benefiting even the most passive users. The method automatically associates, whenever possible, an appropriate semantic tag with a given email. This gives rise to an alternate mechanism for organizing and searching email. We advocate a novel modeling approach that exploits the overall population of users, thereby learning from the wisdom-of-crowds how to categorize messages. Given our massive user base, it is enough to learn from a minority of the users who label certain messages in order to label that kind of messages for the general population. We design a novel cascade classification approach, which copes with the severe scalability and accuracy constraints we are facing. Significant efficiency gains are achieved by working within a low dimensional latent space, and by using a novel hierarchical classifier. Precision level is controlled by separating the task into a two-phase classification process. We performed an extensive empirical study covering three different time periods, over 100 million messages, and thousands of candidate tags per message. The results are encouraging and compare favorably with alternative approaches. Our method successfully tags 72% of incoming email traffic. Performance-wise, the computational overhead, even on surge large traffic, is sufficiently low for our approach to be applicable in production on any large Web mail service.

Recommender systems perform much better on users for which they have more information. This gives rise to a problem of satisfying users new to a system. The problem is even more acute considering that some of these hard to profile new users judge the unfamiliar system by its ability to immediately provide them with satisfying recommendations, and may quickly abandon the system when disappointed. Rapid profiling of new users by a recommender system is often achieved through a bootstrapping process - a kind of an initial interview - that elicits users to provide their opinions on certain carefully chosen items or categories. The elicitation process becomes particularly effective when adapted to users' responses, making best use of users' time by dynamically modifying the questions to improve the evolving profile. In particular, we advocate a specialized version of decision trees as the most appropriate tool for this task. We detail an efficient tree learning algorithm, specifically tailored to the unique properties of the problem. Several extensions to the tree construction are also introduced, which enhance the efficiency and utility of the method. We implemented our methods within a movie recommendation service. The experimental study delivered encouraging results, with the tree-based bootstrapping process significantly outperforming previous approaches.

The collaborative filtering (CF) approach to recommenders has recently enjoyed much interest and progress. The fact that it played a central role within the recently completed Netflix competition has contributed to its popularity. This chapter surveys the recent progress in the field. Matrix factorization techniques, which became a first choice for implementing CF, are described together with recent innovations. We also describe several extensions that bring competitive accuracy into neighborhood methods, which used to dominate the field. The chapter demonstrates how to utilize temporal models and implicit feedback to extend models accuracy. In passing, we include detailed descriptions of some the central methods developed for tackling the challenge of the Netflix Prize competition.

Recommender systems perform much better on users for which they have more information. This gives rise to a problem of satisfying users new to a system. The problem is even more acute considering that some of these hard to profile new users judge the unfamiliar system by its ability to immediately provide them with satisfying recommendations, and may be the quickest to abandon the system when disappointed. Rapid profiling of new users is often achieved through a bootstrapping process - a kind of an initial interview - that elicits users to provide their opinions on certain carefully chosen items or categories. This work offers a new bootstrapping method, which is based on a concrete optimization goal, thereby handily outperforming known approaches in our tests.

In many commercial systems, the 'best bet' recommendations are shown, but the predicted rating values are not. This is usually referred to as a top-N recommendation task, where the goal of the recommender system is to find a few specific items which are supposed to be most appealing to the user. Common methodologies based on error metrics (such as RMSE) are not a natural fit for evaluating the top-N recommendation task. Rather, top-N performance can be directly measured by alternative methodologies based on accuracy metrics (such as precision/recall). An extensive evaluation of several state-of-the art recommender algorithms suggests that algorithms optimized for minimizing RMSE do not necessarily perform as expected in terms of top-N recommendation task. Results show that improvements in RMSE often do not translate into accuracy improvements. In particular, a naive non-personalized algorithm can outperform some common recommendation approaches and almost match the accuracy of sophisticated algorithms. Another finding is that the very few top popular items can skew the top-N performance. The analysis points out that when evaluating a recommender algorithm on the top-N recommendation task, the test set should be chosen carefully in order to not bias accuracy metrics towards non-personalized solutions. Finally, we offer practitioners new variants of two collaborative filtering algorithms that, regardless of their RMSE, significantly outperform other recommender algorithms in pursuing the top-N recommendation task, with offering additional practical advantages. This comes at surprise given the simplicity of these two methods.

In this tutorial we discuss the evaluation of recommender systems. We discuss the main reason for evaluating recommender systems, i.e., the selection task. We overview some general guidelines for conducting evaluation tests. We then discuss the evaluation ...

Customer preferences for products are drifting over time. Product perception and popularity are constantly changing as new selection emerges. Similarly, customer inclinations are evolving, leading them to ever redefine their taste. Thus, modeling temporal dynamics is essential for designing recommender systems or general customer preference models. However, this raises unique challenges. Within the ecosystem intersecting multiple products and customers, many different characteristics are shifting simultaneously, while many of them influence each other and often those shifts are delicate and associated with a few data instances. This distinguishes the problem from concept drift explorations, where mostly a single concept is tracked. Classical time-window or instance decay approaches cannot work, as they lose too many signals when discarding data instances. A more sensitive approach is required, which can make better distinctions between transient effects and long-term patterns. We show how to model the time changing behavior throughout the life span of the data. Such a model allows us to exploit the relevant components of all data instances, while discarding only what is modeled as being irrelevant. Accordingly, we revamp two leading collaborative filtering recommendation approaches. Evaluation is made on a large movie-rating dataset underlying the Netflix Prize contest. Results are encouraging and better than those previously reported on this dataset. In particular, methods described in this paper play a significant role in the solution that won the Netflix contest.