Due to the need to protect personal information and the impracticality of exhaustive data collection, there is increasing need to deal with datasets with various levels of granularity, such as user-individual data and user-group data. In this study, we propose a new method for jointly analyzing multiple datasets with different granularity. The proposed method is a probabilistic model based on nonnegative matrix factorization, which is derived by introducing latent variables that indicate the high-resolution data underlying the low-resolution data. Experiments on purchase logs show that the proposed method has a better performance than the existing methods. Furthermore, by deriving an extension of the proposed method, we show that the proposed method is a new fundamental approach for analyzing datasets with different granularity.
Masahiro KOHJIMA
NTT Service Evolution Laboratories
Tatsushi MATSUBAYASHI
NTT Service Evolution Laboratories
Hiroshi SAWADA
NTT Service Evolution Laboratories
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Masahiro KOHJIMA, Tatsushi MATSUBAYASHI, Hiroshi SAWADA, "Learning of Nonnegative Matrix Factorization Models for Inconsistent Resolution Dataset Analysis" in IEICE TRANSACTIONS on Information,
vol. E102-D, no. 4, pp. 715-723, April 2019, doi: 10.1587/transinf.2018AWI0002.
Abstract: Due to the need to protect personal information and the impracticality of exhaustive data collection, there is increasing need to deal with datasets with various levels of granularity, such as user-individual data and user-group data. In this study, we propose a new method for jointly analyzing multiple datasets with different granularity. The proposed method is a probabilistic model based on nonnegative matrix factorization, which is derived by introducing latent variables that indicate the high-resolution data underlying the low-resolution data. Experiments on purchase logs show that the proposed method has a better performance than the existing methods. Furthermore, by deriving an extension of the proposed method, we show that the proposed method is a new fundamental approach for analyzing datasets with different granularity.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2018AWI0002/_p
Copy
@ARTICLE{e102-d_4_715,
author={Masahiro KOHJIMA, Tatsushi MATSUBAYASHI, Hiroshi SAWADA, },
journal={IEICE TRANSACTIONS on Information},
title={Learning of Nonnegative Matrix Factorization Models for Inconsistent Resolution Dataset Analysis},
year={2019},
volume={E102-D},
number={4},
pages={715-723},
abstract={Due to the need to protect personal information and the impracticality of exhaustive data collection, there is increasing need to deal with datasets with various levels of granularity, such as user-individual data and user-group data. In this study, we propose a new method for jointly analyzing multiple datasets with different granularity. The proposed method is a probabilistic model based on nonnegative matrix factorization, which is derived by introducing latent variables that indicate the high-resolution data underlying the low-resolution data. Experiments on purchase logs show that the proposed method has a better performance than the existing methods. Furthermore, by deriving an extension of the proposed method, we show that the proposed method is a new fundamental approach for analyzing datasets with different granularity.},
keywords={},
doi={10.1587/transinf.2018AWI0002},
ISSN={1745-1361},
month={April},}
Copy
TY - JOUR
TI - Learning of Nonnegative Matrix Factorization Models for Inconsistent Resolution Dataset Analysis
T2 - IEICE TRANSACTIONS on Information
SP - 715
EP - 723
AU - Masahiro KOHJIMA
AU - Tatsushi MATSUBAYASHI
AU - Hiroshi SAWADA
PY - 2019
DO - 10.1587/transinf.2018AWI0002
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E102-D
IS - 4
JA - IEICE TRANSACTIONS on Information
Y1 - April 2019
AB - Due to the need to protect personal information and the impracticality of exhaustive data collection, there is increasing need to deal with datasets with various levels of granularity, such as user-individual data and user-group data. In this study, we propose a new method for jointly analyzing multiple datasets with different granularity. The proposed method is a probabilistic model based on nonnegative matrix factorization, which is derived by introducing latent variables that indicate the high-resolution data underlying the low-resolution data. Experiments on purchase logs show that the proposed method has a better performance than the existing methods. Furthermore, by deriving an extension of the proposed method, we show that the proposed method is a new fundamental approach for analyzing datasets with different granularity.
ER -