2018
DOI: 10.1007/s10115-018-1197-7
|View full text |Cite
|
Sign up to set email alerts
|

EigenRec: generalizing PureSVD for effective and efficient top-N recommendations

Abstract: We introduce E R ; a versatile and efficient Latent-Factor framework for Top-N Recommendations that includes the well-known PureSVD algorithm as a special case. E R builds a low dimensional model of an inter-item proximity matrix that combines a similarity component, with a scaling operator, designed to control the influence of the prior item popularity on the final model. Seeing PureSVD within our framework provides intuition about its inner workings, exposes its inherent limitations, and also, paves the path… Show more

Help me understand this report
View preprint versions

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

0
21
0

Year Published

2019
2019
2023
2023

Publication Types

Select...
6
1
1

Relationship

1
7

Authors

Journals

citations
Cited by 34 publications
(21 citation statements)
references
References 35 publications
0
21
0
Order By: Relevance
“…To evaluate the top-n recommendation performance, we adopted the widely used leave-one-out evaluation protocol [8,12,21,26]. In particular, for each user we randomly select one liked 3 item and we create a test set T .…”
Section: Evaluation Methodology and Metricsmentioning
confidence: 99%
See 1 more Smart Citation
“…To evaluate the top-n recommendation performance, we adopted the widely used leave-one-out evaluation protocol [8,12,21,26]. In particular, for each user we randomly select one liked 3 item and we create a test set T .…”
Section: Evaluation Methodology and Metricsmentioning
confidence: 99%
“…Over the past decade a vast number of algorithms have been proposed to tackle the top-n recommendation task. These include neighborhood-based methods [15,23,24]; latent-space methods [8,13,18,21,29]; graph-based methods [5,7,9,10,16,22]; and more recently methods relying on deep neural networks [12,19,30]. PerDif brings together item-models with random walks, and thus lies at the intersection between neighborhood-and graph-based methods; the item transition component captures neighborhood information of the items which is then integrated in a randomwalk-based framework to learn personalized difusions for top-n recommendations.…”
Section: Related Workmentioning
confidence: 99%
“…Due to the information explosion, recommender systems (RS) have become an important tool for people to find the information they need in terms of recommending queries [1]- [3] or items [4]. Most traditional approaches are derived from collaborative filtering [5] or matrix factorization [6]- [9], and give a recommendation list mainly based on a so-called user rating matrix which records users' explicit or implicit interactions with items. However, the interaction in the form of a static user rating matrix is limited because the matrix is time-independent.…”
Section: Introductionmentioning
confidence: 99%
“…In recent years, recommender systems which can only retain relevant information have received increasing attention [1][2][3]. Researchers proposed many methods to solve the recommendation task, which can be classified into four classes: neighborhood-based methods [4][5][6][7], model-based methods [8][9][10][11], graph-based methods [12][13][14][15][16], and deep neural network based methods [17][18][19][20][21]. Neighborhoodbased methods contain user-based collaborative filtering and item-based collaborative filtering, this kind of methods utilize neighbor information of users to make prediction [22].…”
Section: Introductionmentioning
confidence: 99%