SELF TRANSFER LEARNING RECOMMENDATION METHOD AND SYSTEM

A computer implemented method for recommending an item to a user in big data systems, according to which a binary matrix of ratings of items is generated by users for a full target domain and sub-matrixes with high correlation between the users, items and their ratings are discovered in the binary m...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Hauptverfasser:	DAYAN, AVIRAM, INY, YONI, SHAPIRA, BRACHA, GROLMAN, EDITA, BAR, ARIEL, ROKACH, LIOR
Format:	Patent
Sprache:	eng ; fre ; ger
Schlagworte:	CALCULATING COMPUTING COUNTING DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES PHYSICS SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR
Online-Zugang:	Volltext bestellen
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Beschreibung
Zusammenfassung:	A computer implemented method for recommending an item to a user in big data systems, according to which a binary matrix of ratings of items is generated by users for a full target domain and sub-matrixes with high correlation between the users, items and their ratings are discovered in the binary matrix. The density of the sub-matrixes in the binary matrix is calculated and a source domain matrix is populated according to the results, based on the density calculated in the discovered sub-matrixes. Then top-N matrixes with highest densities are selected. Codebooks, which contains different user-item rating patterns, are constructed for each of the selected sub-matrixes, which are sub-domains and constructing a codebook for the matrix with the full rating dataset. Finally the captured different user-item rating patterns are projected to the full domain dataset, based on the constructed codebooks.