Facebook

About: Facebook is a company organization based out in Tel Aviv, Israel. It is known for research contribution in the topics: Computer science & Artificial neural network. The organization has 7856 authors who have published 10906 publications receiving 570123 citations. The organization is also known as: facebook.com & FB.

Embedding-based Retrieval in Facebook Search

Abstract: Search in social networks such as Facebook poses different challenges than in classical web search: besides the query text, it is important to take into account the searcher's context to provide relevant results Their social graph is an integral part of this context and is a unique aspect of Facebook search While embedding-based retrieval (EBR) has been applied in eb search engines for years, Facebook search was still mainly based on a Boolean matching model In this paper, we discuss the techniques for applying EBR to a Facebook Search system We introduce the unified embedding framework developed to model semantic embeddings for personalized search, and the system to serve embedding-based retrieval in a typical search system based on an inverted index We discuss various tricks and experiences on end-to-end optimization of the whole system, including ANN parameter tuning and full-stack optimization Finally, we present our progress on two selected advanced topics about modeling We evaluated EBR on verticals for Facebook Search with significant metrics gains observed in online A/B experiments We believe this paper will provide useful insights and experiences to help people on developing embedding-based retrieval systems in search engines

Modeling Facial Geometry Using Compositional VAEs

Abstract: We propose a method for learning non-linear face geometry representations using deep generative models. Our model is a variational autoencoder with multiple levels of hidden variables where lower layers capture global geometry and higher ones encode more local deformations. Based on that, we propose a new parameterization of facial geometry that naturally decomposes the structure of the human face into a set of semantically meaningful levels of detail. This parameterization enables us to do model fitting while capturing varying level of detail under different types of geometrical constraints.

Method and system for managing digital assets

Abstract: The claimed invention enables digital asset management that is responsive to a user's interactions with digital assets (32). Based on the user's interaction, the invention generates a ranking of the digital assets that is intended to reflect the value of the digital assets to the user. The ranking (36) is based in part on the access frequency and recency, and the number and types of uses of the digital assets. An access hierarchy is derived from the ranking that stores the digital assets so that the higher ranked digital assets are more easily accessed than the lower ranked digital assets. The digital assets can include any of digital images, audio files, and Uniform Resource Locators. The invention can also distinguish between different types of uses so that some types of uses imbue an asset with greater value than others. The types of uses include passive viewing or playback, file sharing, transport, and editing. The invention also allows the user to assign subjective values to each digital asset that can be factored into the ranking independently of usage patterns. A volatility-dampening attribute is provided to moderate volatility in the access hierarchy.

Understanding Back-Translation at Scale

Abstract: An effective method to improve neural machine translation with monolingual data is to augment the parallel training corpus with back-translations of target language sentences. This work broadens the understanding of back-translation and investigates a number of methods to generate synthetic source sentences. We find that in all but resource poor settings back-translations obtained via sampling or noised beam outputs are most effective. Our analysis shows that sampling or noisy synthetic data gives a much stronger training signal than data generated by beam or greedy search. We also compare how synthetic data compares to genuine bitext and study various domain effects. Finally, we scale to hundreds of millions of monolingual sentences and achieve a new state of the art of 35 BLEU on the WMT'14 English-German test set.

Nonparametric Link Prediction in Dynamic Networks

Abstract: We propose a non-parametric link prediction algorithm for a sequence of graph snapshots over time. The model predicts links based on the features of its endpoints, as well as those of the local neighborhood around the endpoints. This allows for different types of neighborhoods in a graph, each with its own dynamics (e.g, growing or shrinking communities). We prove the consistency of our estimator, and give a fast implementation based on locality-sensitive hashing. Experiments with simulated as well as five real-world dynamic graphs show that we outperform the state of the art, especially when sharp fluctuations or non-linearities are present.