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

Item-based models are among the most popular collaborative filteringapproaches for building recommender systems. Random walks can provide apowerful tool for harvesting the rich network of interactions captured withinthese models. They can exploit indirect relations between the items, mitigatethe effects of sparsity, ensure wider itemspace coverage, as well as increasethe diversity of recommendation lists. Their potential, however, can behindered by the tendency of the walks to rapidly concentrate towards thecentral nodes of the graph, thereby significantly restricting the range ofK-step distributions that can be exploited for personalized recommendations. Inthis work we introduce RecWalk; a novel random walk-based method that leveragesthe spectral properties of nearly uncoupled Markov chains to provably lift thislimitation and prolong the influence of users' past preferences on thesuccessive steps of the walk---allowing the walker to explore the underlyingnetwork more fruitfully. A comprehensive set of experiments on real-worlddatasets verify the theoretically predicted properties of the proposed approachand indicate that they are directly linked to significant improvements in top-nrecommendation accuracy. They also highlight RecWalk's potential in providing aframework for boosting the performance of item-based models. RecWalk achievesstate-of-the-art top-n recommendation quality outperforming several competingapproaches, including recently proposed methods that rely on deep neuralnetworks.

Boosting Item-based Collaborative Filtering via Nearly Uncoupled Random Walks