Modeling the Effects of Individual and Group Heterogeneity on Multi-Aspect Rating Behavior

doi:10.11871/jfdc.issn.2096-742X.2020.02.005

Abstract

Abstract:

[Objective] Multi-aspect rating system could help customers better understand the item or service, because it provides not only the overall rating but also more detailed aspect ratings. By modeling the rating patterns on multi-aspect rating systems, we can better find out latent rating groups and quantitatively understand the rating behaviors lie in these groups. This can also help service providers improve their service and attract more targeted customers. However, due to the complex nature of multi-aspect rating system, it is challenging to model its rating patterns. [Methods] To address this problem, in this paper, we propose a two-step framework to learn the rating patterns from multi-aspect rating systems. Specifically, we first propose a multi-factorization relationship learning (MFRL) method to obtain the user and item aspect factor matrices. In MFRL, we unify matrix factorization, multi-task learning and task relationship learning into one optimization framework. And then, we model the rating patterns by exploiting group-wise overall rating prediction via mixture regression, whose inputs are the factor vectors of users and items learned from MFRL method. [Results] We apply the proposed framework on a real-world dataset (i.e., the crawled hotel rating dataset from TripAdvisor.com) to evaluate the performance of our proposed method. Extensive experimental results demonstrate the effectiveness of the proposed framework. [Conclusions] Individual and Group Heterogeneity could affect the behaviors behind the rating acts, which should be taken into account in modeling the rating patterns.

Key words: Multi-Aspect Rating, Recommender System, Multi-Task Learning, Relationship Learning, User Behavior

Liu Kunpeng,Zhao Xiaosa,Hu Yirui,Fu Yanjie. Modeling the Effects of Individual and Group Heterogeneity on Multi-Aspect Rating Behavior[J]. Frontiers of Data and Computing, 2020, 2(2): 59-77.

Figures/Tables 23

1.	Generate latent factors Generate ${{u}_{i}}$, ${{e}_{j}}$, ${{c}_{j}}$ from MFRL
2.	Generate distributions for each group a. Draw ${{\lambda }_{n}}$, $\sigma _{n}^{2}$ for each group b. Draw ${{f}_{n}}(x)\tilde{\ }N([u_{i}^{T},e_{j}^{T},c_{j}^{T}]\cdot {{\lambda }_{n}},\sigma _{n}^{2}$
3.	Generate overall rating b. Generate weight in the mixture model $α n$ c. Generate ${{y}_{ij}}\tilde{\ }\sum\nolimits_{n=1}^{N}{{{\alpha }_{n}}N([u_{i}^{T},e_{j}^{T},c_{j}^{T}]\cdot {{\lambda }_{n}},\sigma _{n}^{2})}$

References 43

[1]	McAuley J, Leskovec J & Jurafsky D . Learning attitudes and attributes from multi-aspect reviews [C]. In Data Mining (ICDM), 2012 IEEE 12th International Conference on. IEEE 2012 (2012):1020-1025.
[2]	Fu X H, Liu G, Guo Y Y et al. Multi-aspect sentiment analysis for Chinese online social reviews based on topic modeling and HowNet lexicon[J]. Knowledge-Based Systems 37(2013):186-195.
[3]	Zhu J , Zhang C & Ma M Y . Multi-aspect rating inference with aspect-based segmentation[J]. IEEE Transactions on Affective Computing 3(2012):469-481.
[4]	Baas SM & Kwakernaak H . Rating and ranking of multiple-aspect alternatives using fuzzy sets[J]. Automatica 13(1977):47-58.
[5]	Bin Lu B, Myle Ott M, Claire Cardie C, et al. Multi-aspect sentiment analysis with topic models [C]. In Data Mining Workshops (ICDMW), 2011 IEEE 11th International Conference on. IEEE 2011 (2011):81-88.
[6]	Fang Y & Si L . Matrix co-factorization for recommendation with rich side information and implicit feedback [C]. In Proceedings of the 2nd International Workshop on Information Heterogeneity and Fusion in Recommender Systems. ACM 2011 (2011):65-69.
[7]	Klami A, Bouchard G & Tripathi A. Group-sparse embeddings in collective matrix factorization[J]. arXiv preprint arXiv:1312.5921, 2013.
[8]	Marlin B & Zemel R S . The multiple multiplicative factor model for collaborative filtering [C]. In Proceedings of the twenty-first international conference on Machine learning. ACM 2004 (2004):73.
[9]	Mnih A & Salakhutdinov R R . Probabilistic matrix factorization[J]. In Advances in neural information processing systems 2008 (2008):1257-1264.
[10]	Singh A P & Gordon G J . Relational learning via collective matrix factorization [C]. In Proceedings of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining. ACM 2008 ( 2008):650-658.
[11]	Schmidt M N, Winther O & Hansen L K. Bayesian non-negative matrix factorization [C]. In International Conference on Independent Component Analysis and Signal Separation. Springer 2009 (2009):540-547.
[12]	Moon T K . 1996. The expectation-maximization algorithm[J]. IEEE Signal processing magazine 13(1996):47-60.
[13]	Linoff G S & Berry M J A . Data mining techniques: for marketing, sales, and customer relationship management[J]. John Wiley & Sons, 2011.
[14]	Fu Y J, Liu B, Ge Y et al. User preference learning with multiple information fusion for restaurant recommendation [C]. In Proceedings of the 2014 SIAM International Conference on Data Mining. SIAM 2014 (2014):470-478.
[15]	Zhou H Y, Chen J H & Ye J P. Malsar: Multi-task learning via structural regularization[J]. Arizona State University 21 (2011).
[16]	Vallerand R J. Toward a hierarchical model of intrinsic and extrinsic motivation. In Advances in experimental social psychology[J]. Academic Press 29 (1997):271-360.
[17]	Lee H H & Teng W G . Incorporating multi-criteria ratings in recommendation systems [C]. In Information Reuse and Integration, 2007. IRI 2007. IEEE International Conference on. IEEE 2007 (2007):273-278.
[18]	Adomavicius G, Sankaranarayanan R, Sen S et al. Incorporating contextual information in recommender systems using a multidimensional approach[J]. ACM Transactions on Information Systems 23(2005):103-145.
[19]	Adomavicius G & Kwon Y O. 2007. New recommendation techniques for multicriteria rating systems[J]. IEEE Intelligent Systems 22(2007):48-55.
[20]	Argyriou A, Evgeniou T & Pontil M. Multi-task feature learning[J]. In Advances in neural information processing systems 2007 (2007):41-48.
[21]	Jalali A, Sanghavi S, Ruan C & Ravikumar P K. . A dirty model for multi-task learning[J]. In Advances in neural information processing systems 2010 (2010):964-972.
[22]	Lounici K, Pontil M, Tsybakov A B , et al. Taking advantage of sparsity in multi-task learning[J]. arXiv preprint arXiv:0903.1468, 2009.
[23]	Zhou J Y, Yuan L, Liu J et al. A multi-task learning formulation for predicting disease progression [C]. In Proceedings of the 17th ACM SIGKDD international conference on Knowledge discovery and data mining. ACM 2011 (2011):814-822.
[24]	Kumar A & Daume III H . Learning task grouping and overlap in multi-task learning[J]. arXiv preprint arXiv:1206.6417, 2012.
[25]	Evgeniou T & Pontil M . Regularized multi-task learning [C]. In Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining. ACM 2004 (2004):109-117.
[26]	Liu J, Ji S W & Ye J P . Multi-task feature learning via efficient l 2, 1-norm minimization[C]. In Proceedings of the twenty-fifth conference on uncertainty in artificial intelligence. AUAI Press 2009 (2009):339-348.
[27]	Zhang Y & Yeung D Y . A convex formulation for learning task relationships in multi-task learning[J]. arXiv preprint arXiv:1203.3536, 2012.
[28]	Xue Y, Liao X J, Carin L et al. Multi-task learning for classification with dirichlet process priors[J]. Journal of Machine Learning Research 8(2007):35-63.
[29]	Figueiredo M A T & Jain A K . Unsupervised learning of finite mixture models[J]. IEEE Transactions on pattern analysis and machine intelligence 24(2002):381-396.
[30]	Zivkovic Z & Heijden F V D . Recursive unsupervised learning of finite mixture models[J]. IEEE Transactions on pattern analysis and machine intelligence 26(2004):651-656.
[31]	Chen H F, Chen J H & Kalbfleisch J D . A modified likelihood ratio test for homogeneity in finite mixture models[J]. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 63(2001):19-29.
[32]	Bapna R, Goes G, Wei K K et al. A finite mixture logit model to segment and predict electronic payments system adoption[J]. Information Systems Research 22(2011):118-133.
[33]	Muthén B & Shedden K . Finite mixture modeling with mixture outcomes using the EM algorithm[J]. Biometrics 55(1999):463-469.
[34]	Yu J & Qin S J . Multimode process monitoring with Bayesian inference-based finite Gaussian mixture models[J]. AIChE Journal 54(2008):1811-1829.
[35]	Teratanavat R and Hooker N H . Consumer valuations and preference heterogeneity for a novel functional food[J]. Journal of Food Science 71(2006):S533-S541.
[36]	Cicia G, Del Giudice T & Scarpa R . Consumers’ perception of quality in organic food: a random utility model under preference heterogeneity and choice correlation from rank-orderings[J]. British Food Journal 104(2002):200-213.
[37]	Ravi Dhar . 1997. Consumer preference for a no-choice option[J]. Journal of consumer research 24, 2(1997), 215-231.
[38]	Kamakura W A, Kim B D & Lee J . 1996. Modeling preference and structural heterogeneity in consumer choice[J]. Marketing Science 15(1996):152-172.
[39]	Liu K., Wang P., Zhang J., Fu Y. and Das S.K., 2018, May. Modeling the Interaction Coupling of Multi-View Spatiotemporal Contexts for Destination Prediction [C]. In Proceedings of the 2018 SIAM International Conference on Data Mining (pp. 171-179). Society for Industrial and Applied Mathematics.
[40]	Liu K., Uplavikar N., Jiang W. and Fu Y., 2018, November. Privacy-Preserving Multi-task Learning [C]. In 2018 IEEE International Conference on Data Mining (ICDM)(pp. 1128-1133). IEEE.
[41]	Liu K., Fu Y., Wang P., Wu L., Bo R. and Li X., 2019, July. Automating Feature Subspace Exploration via Multi-Agent Reinforcement Learning [C]. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (pp. 207-215).
[42]	Breffle W S & Morey E R . Investigating preference heterogeneity in a repeated discrete-choice recreation demand model of Atlantic salmon fishing[J]. Marine Resource Economics 15(2000):1-20.
[43]	Poulsen C S, Brockhoff P M B & Erichsen L . Heterogeneity in consumer preference data-a combined approach[J]. Food quality and preference 8 5-6(1997):409-417.

Symbol	Size	Description
Y	I×J	Overall rating matrix
G	I×J	Aspect-2 rating matrix
H	I×J	Aspect-1 rating matrix
y_ij	1	Overall rating of user i for item j
g_ij	1	Aspect-1 rating of user i for item j
h_ij	1	aspect-2 rating of user i for item j
U	K×1	User latent matrix
E	K×J	Item aspect-1 latent matrix
C	K×J	Item aspect-2 latent matrix
u_i	K×1	Latent features of user i
e_j	K×1	Aspect-1 latent features of item j
c_j	K×1	Aspect-2 latent features of item j
u_ki	1	k-th latent feature of user i
e_kj	1	k-th latent feature of aspect-1 of item j
c_kj	1	k-th latent feature of aspect-2 of item j

Overall Rating	${{y}_{ij}}\tilde{\ }N(\mu _{ij}^{y},{{\sigma }^{2}})$
Aspect-1 Rating	${{g}_{ij}}\tilde{\ }N(\mu _{ij}^{g},{{\sigma }^{2}})$
Aspect-2 Rating	${{h}_{ij}}\tilde{\ }N(\mu _{ij}^{h},{{\sigma }^{2}})$
Overall Utility	$\mu _{ij}^{y}=p\cdot u_{i}^{T}{{e}_{j}}+q\cdot u_{i}^{T}{{c}_{j}}$
Aspect-1 Utility	$\mu _{ij}^{g}=u_{i}^{T}{{e}_{j}}$
Aspect-2 Utility	$\mu _{ij}^{h}=u_{i}^{T}{{c}_{j}}$
User Latent Factors	${{u}_{ki}}\tilde{\ }Exp(\alpha )$
Aspect-1 Latent Factors	${{e}_{kj}}\tilde{\ }Exp(\beta )$
Aspect-2 Latent Factors	${{c}_{kj}}\tilde{\ }Exp(\beta )$
Variance	${{\sigma }^{2}}\tilde{\ }Inv-Gamma(a,b)$

Methods	K=5					K=20
	MAE	RMSE	NDCG@1	NDCG@3	NDCG@5	MAE	RMSE	NDCG@1	NDCG@3	NDCG@5
MFRL(R1)	0.907	1.100	0.721	0.751	0.702	0.915	1.117	0.784	0.732	0.705
MFRL(R2)	0.912	1.073	0.706	0.784	0.697	0.881	1.251	0.799	0.738	0.711
MFRL(R3)	0.903	1.106	0.732	0.770	0.707	0.914	1.228	0.807	0.727	0.691
MFRL(R4)	0.915	1.119	0.710	0.783	0.716	0.901	1.255	0.791	0.742	0.709
MFRL(R5)	0.912	1.092	0.714	0.762	0.710	0.925	1.239	0.810	0.715	0.699
Mix-MFRL(R1)	0.792	0.970	0.771	0.831	0.724	0.813	1.009	0.791	0.801	0.726
Mix-MFRL(R2)	0.813	0.995	0.769	0.812	0.742	0.837	1.052	0.779	0.809	0.744
Mix-MFRL(R3)	0.782	0.930	0.757	0.794	0.734	0.797	0.819	0.787	0.813	0.709
Mix-MFRL(R4)	0.801	1.014	0.780	0.805	0.781	0.781	1.280	0.799	0.786	0.727
Mix-MFRL(R5)	0.810	0.978	0.746	0.830	0.762	0.774	1.212	0.781	0.791	0.736