/Text2Path

[NAACL(2019)] Generating Knowledge Graph Paths from Textual Definitions using Sequence-to-Sequence Models

Primary LanguageShell

Text2Path: Model for Mapping Unrestricted Text to Knowledge Graph Entities

Table of contents

  1. Replicate Results
  2. Usage
  3. Citing
  4. Licence
  5. Contact info

Replicate Results

We provide pretrained models (all models were trained on a CPU) for text2edges(*)[last row of Table 2]. To replicate the results:

Step 1:

$ cd ./Text2ath/Scripts/Model/Trained_Models/

Unzip the pretrained models in the directory.

Step 2:

$ cd ./Text2Path/Scripts/Model/
$ sh get_reported_results.sh

Code for the MS-LSTM model can be found here

Usage

Step 1: Graph Preprocessing

Algorithm assumes that a graph is a rooted tree and it is represented as an edge list:

node1 node5
node2 node4
...

Furthermore each node in the graph must have a textual definition:

node1 text_def_1
node2 text_def_2
...

To preprocess a graph for a text2nodes model:

$ cd ./Text2Path/Scripts/Preprocessing/
$ sh make_path_node_representation_dataset.sh

To preprocess a graph for a text2edges model:

$ cd ./Text2Path/Scripts/Preprocessing/
$ sh make_artificial_vocab_representation_dataset.sh

One can get the pretrained word embeddings used in the experiments here

Step2 Train a Model:

To train a new model:

$ cd ./Text2Path/Scripts/Model/
$ python text_to_path_model.py --train_data <path_to_train_data> --augment_data <augment_data_file> --test_data <test_data_file> --checkpoint <save_model_file> --graph <graph_file> --is_train 1

Citing

If you find this material useful in your research, please cite:

@InProceedings{prokhorov_etal:NAACL2019,
  author={Victor Prokhorov and Mohammad T. Pilehvar and Nigel Collier},
  title={Generating Knowledge Graph Paths from Textual Definitions using Sequence-to-Sequence Models},
  booktitle={Proceedings of the 2019 Annual Conference of the North American Chapter of the Association for Computational Linguistics},
  year={2019},
  month={June},
  address={Minneapolis, USA},
  publisher={Association for {C}omputational {L}inguistics}
}

Licence

The code in this repository is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License version 3 as published by the Free Software Foundation. The code is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

Contact info

For questions or more information please use the following: