2008 International Conference on Automated Solutions for Cross Media Content and Multi-Channel Distribution 2008
DOI: 10.1109/axmedis.2008.33
|View full text |Cite
|
Sign up to set email alerts
|

Experimental Results on Item-Based Algorithms for Independent Domain Collaborative Filtering

Abstract: A research analysis on item-based algorithms for collaborative filtering is presented. The aim of the presented activity was to find a configuration of an item-based algorithm capable of providing good results but also independent from the data set. Four data sets were used for the algorithm validation: Netflix, MovieLens, BookCrossing, and Jester. The experimentation involved the following aspects: similarity computation, size of the neighbourhood, prediction computation, minimum number of co-rated items. Res… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

0
2
0
1

Year Published

2009
2009
2016
2016

Publication Types

Select...
5
2

Relationship

1
6

Authors

Journals

citations
Cited by 7 publications
(3 citation statements)
references
References 11 publications
0
2
0
1
Order By: Relevance
“…This data has been cleaned up -users who had less than 20 ratings or did not have complete demographic information were removed from this data set. This dataset are being put into use on the field of collaborative filtering [19] [20] , recommender system [21] [22] and so on.…”
Section: Data Setmentioning
confidence: 99%
“…This data has been cleaned up -users who had less than 20 ratings or did not have complete demographic information were removed from this data set. This dataset are being put into use on the field of collaborative filtering [19] [20] , recommender system [21] [22] and so on.…”
Section: Data Setmentioning
confidence: 99%
“…While before the Netflix competition the itembased algorithms were considered the most effective for recommender systems, and in fact at the time they were used also by Amazon (Linden, Smith & York, 2003;Clemente, 2008), during the competition it has been demonstrated that the matrix factorization algorithms, working alone, were the most effective for this kind of problems (Koren, Bell & Volinsky, 2009;Tosher, Jahrer & Bell, 2009).…”
Section: Related Workmentioning
confidence: 99%
“…Επιπλέον, η χρησιμοποίηση ενός κατασκευαστικού αλγορίθμου δικτύου φαίνεται ότι αποτελεί ένα ακόμα πλεονέκτημα, Xu and Yang, 2011;Mackey et al, 2010] εμφανίζουν ένα μικρό, αλλά σταθερό, προβάδισμα σε σύγκριση με το ksepRS. Παρόλα αυτά, η προτεινόμενη μεθοδολογία καταφέρνει να είναι περισσότερο ανταγωνιστική σε σύγκριση με τη βασική μέθοδο ή άλλες γενικές μεθόδους συνεργατικής διήθησης, απλές [Desrosiers and Karypis, 2010] και βασιζόμενες στα αντικείμενα [Clemente, 2008], οι οποίες φαίνεται ότι δεν μπορούν να επεξεργαστούν την εγγενή μη-γραμμικότητα του πίνακα αξιολογήσεων των χρηστών.…”
Section: πειραματικό πρωτόκολλοunclassified