English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Collaborative filtering algorithms are important building blocks in manypractical recommendation systems. For example, many large-scale data processingenvironments include collaborative filtering models for which the AlternatingLeast Squares (ALS) algorithm is used to compute latent factor matrixdecompositions. In this paper, we propose an approach to accelerate theconvergence of parallel ALS-based optimization methods for collaborativefiltering using a nonlinear conjugate gradient (NCG) wrapper around the ALSiterations. We also provide a parallel implementation of the acceleratedALS-NCG algorithm in the Apache Spark distributed data processing environment,and an efficient line search technique as part of the ALS-NCG implementationthat requires only one pass over the data on distributed datasets. In serialnumerical experiments on a linux workstation and parallel numerical experimentson a 16 node cluster with 256 computing cores, we demonstrate that the combinedALS-NCG method requires many fewer iterations and less time than standalone ALSto reach movie rankings with high accuracy on the MovieLens 20M dataset. Inparallel, ALS-NCG can achieve an acceleration factor of 4 or greater in clocktime when an accurate solution is desired; furthermore, the acceleration factorincreases as greater numerical precision is required in the solution. Inaddition, the NCG acceleration mechanism is efficient in parallel and scaleslinearly with problem size on synthetic datasets with up to nearly 1 billionratings. The acceleration mechanism is general and may also be applicable toother optimization methods for collaborative filtering.

Algorithmic Acceleration of Parallel ALS for Collaborative Filtering: Speeding up Distributed Big Data Recommendation in Spark