The official Pytorch implementation of PC-RNN proposed in paper "Patent Citation Dynamics Modeling via Multi-Attention Recurrent Networks", IJCAI 2019.
- Install
Pytorch(tested on >= 1.0.0), please refer to the official website for further details. - Install
numpy
Multiple hyperparameters (e.g., dimension of hidden state) are tunable in the model, see train.py for details. Run bash run.sh for a quickstart.
Title: Patent Citation Dynamics Modeling via Multi-Attention Recurrent Networks
Abstract: Modeling and forecasting forward citations to a patent is a central task for the discovery of emerging technologies and for measuring the pulse of inventive progress. Conventional methods for forecasting these forward citations cast the problem as analysis of temporal point processes which rely on the conditional intensity of previously received citations. Recent approaches model the conditional intensity as a chain of recurrent neural networks to capture memory dependency in hopes of reducing the restrictions of the parametric form of the intensity function. For the problem of patent citations, we observe that forecasting a patent’s chain of citations benefits from not only the patent’s history itself but also from the historical citations of assignees and inventors associated with that patent. In this paper, we propose a sequence-to-sequence model which employs an attention-of-attention mechanism to capture the dependencies of these multiple time sequences. Furthermore, the proposed model is able to forecast both the timestamp and the category of a patent’s next citation. Extensive experiments on a large patent citation dataset collected from USPTO demonstrate that the proposed model outperforms state-of-the-art models at forward citation forecasting.
If you used data or codes in this repo in your research, please cite "Patent Citation Dynamics Modeling via Multi-Attention Recurrent Networks".
@inproceedings{DBLP:conf/ijcai/JiCSFLR19,
author = {Taoran Ji and
Zhiqian Chen and
Nathan Self and
Kaiqun Fu and
Chang{-}Tien Lu and
Naren Ramakrishnan},
title = {Patent Citation Dynamics Modeling via Multi-Attention Recurrent Networks},
booktitle = {Proceedings of the Twenty-Eighth International Joint Conference on
Artificial Intelligence, {IJCAI} 2019, Macao, China, August 10-16,
2019},
pages = {2621--2627},
year = {2019},
crossref = {DBLP:conf/ijcai/2019},
url = {https://doi.org/10.24963/ijcai.2019/364},
doi = {10.24963/ijcai.2019/364},
timestamp = {Tue, 20 Aug 2019 16:18:18 +0200},
biburl = {https://dblp.org/rec/bib/conf/ijcai/JiCSFLR19},
bibsource = {dblp computer science bibliography, https://dblp.org}
}Codes in this repo released under MIT license.
Feel free to contact me (see my profile for email address) if you're interested in related research topics or want to use patent-related data.