Nested Chinese Restaurant Franchise Process: Applications to User Tracking and Document Modeling

Amr Ahmed, Liangjie Hong, Alexander Smola
Proceedings of the 30th International Conference on Machine Learning, PMLR 28(3):1426-1434, 2013.

Abstract

Much natural data is hierarchical in nature. Moreover, this hierarchy is often shared between different instances. We introduce the nested Chinese Restaurant Franchise Process as a means to obtain both hierarchical tree-structured representations for objects, akin to (but more general than) the nested Chinese Restaurant Process while sharing their structure akin to the Hierarchical Dirichlet Process. Moreover, by decoupling the \emphstructure generating part of the process from the components responsible for the observations, we are able to apply the same statistical approach to a variety of user generated data. In particular, we model the joint distribution of microblogs and locations for Twitter for users. This leads to a 40% reduction in location uncertainty relative to the best previously published results. Moreover, we model documents from the NIPS papers dataset, obtaining excellent perplexity relative to (hierarchical) Pachinko allocation and LDA.

Cite this Paper


BibTeX
@InProceedings{pmlr-v28-ahmed13, title = {Nested Chinese Restaurant Franchise Process: Applications to User Tracking and Document Modeling}, author = {Ahmed, Amr and Hong, Liangjie and Smola, Alexander}, booktitle = {Proceedings of the 30th International Conference on Machine Learning}, pages = {1426--1434}, year = {2013}, editor = {Dasgupta, Sanjoy and McAllester, David}, volume = {28}, number = {3}, series = {Proceedings of Machine Learning Research}, address = {Atlanta, Georgia, USA}, month = {17--19 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v28/ahmed13.pdf}, url = {https://proceedings.mlr.press/v28/ahmed13.html}, abstract = {Much natural data is hierarchical in nature. Moreover, this hierarchy is often shared between different instances. We introduce the nested Chinese Restaurant Franchise Process as a means to obtain both hierarchical tree-structured representations for objects, akin to (but more general than) the nested Chinese Restaurant Process while sharing their structure akin to the Hierarchical Dirichlet Process. Moreover, by decoupling the \emphstructure generating part of the process from the components responsible for the observations, we are able to apply the same statistical approach to a variety of user generated data. In particular, we model the joint distribution of microblogs and locations for Twitter for users. This leads to a 40% reduction in location uncertainty relative to the best previously published results. Moreover, we model documents from the NIPS papers dataset, obtaining excellent perplexity relative to (hierarchical) Pachinko allocation and LDA.} }
Endnote
%0 Conference Paper %T Nested Chinese Restaurant Franchise Process: Applications to User Tracking and Document Modeling %A Amr Ahmed %A Liangjie Hong %A Alexander Smola %B Proceedings of the 30th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2013 %E Sanjoy Dasgupta %E David McAllester %F pmlr-v28-ahmed13 %I PMLR %P 1426--1434 %U https://proceedings.mlr.press/v28/ahmed13.html %V 28 %N 3 %X Much natural data is hierarchical in nature. Moreover, this hierarchy is often shared between different instances. We introduce the nested Chinese Restaurant Franchise Process as a means to obtain both hierarchical tree-structured representations for objects, akin to (but more general than) the nested Chinese Restaurant Process while sharing their structure akin to the Hierarchical Dirichlet Process. Moreover, by decoupling the \emphstructure generating part of the process from the components responsible for the observations, we are able to apply the same statistical approach to a variety of user generated data. In particular, we model the joint distribution of microblogs and locations for Twitter for users. This leads to a 40% reduction in location uncertainty relative to the best previously published results. Moreover, we model documents from the NIPS papers dataset, obtaining excellent perplexity relative to (hierarchical) Pachinko allocation and LDA.
RIS
TY - CPAPER TI - Nested Chinese Restaurant Franchise Process: Applications to User Tracking and Document Modeling AU - Amr Ahmed AU - Liangjie Hong AU - Alexander Smola BT - Proceedings of the 30th International Conference on Machine Learning DA - 2013/05/26 ED - Sanjoy Dasgupta ED - David McAllester ID - pmlr-v28-ahmed13 PB - PMLR DP - Proceedings of Machine Learning Research VL - 28 IS - 3 SP - 1426 EP - 1434 L1 - http://proceedings.mlr.press/v28/ahmed13.pdf UR - https://proceedings.mlr.press/v28/ahmed13.html AB - Much natural data is hierarchical in nature. Moreover, this hierarchy is often shared between different instances. We introduce the nested Chinese Restaurant Franchise Process as a means to obtain both hierarchical tree-structured representations for objects, akin to (but more general than) the nested Chinese Restaurant Process while sharing their structure akin to the Hierarchical Dirichlet Process. Moreover, by decoupling the \emphstructure generating part of the process from the components responsible for the observations, we are able to apply the same statistical approach to a variety of user generated data. In particular, we model the joint distribution of microblogs and locations for Twitter for users. This leads to a 40% reduction in location uncertainty relative to the best previously published results. Moreover, we model documents from the NIPS papers dataset, obtaining excellent perplexity relative to (hierarchical) Pachinko allocation and LDA. ER -
APA
Ahmed, A., Hong, L. & Smola, A.. (2013). Nested Chinese Restaurant Franchise Process: Applications to User Tracking and Document Modeling. Proceedings of the 30th International Conference on Machine Learning, in Proceedings of Machine Learning Research 28(3):1426-1434 Available from https://proceedings.mlr.press/v28/ahmed13.html.

Related Material