# embedding-disruptiveness
`embedding-disruptiveness` is a Python package for calculating the Embedding Disruptiveness Index of papers or patents using a citation network. This measure helps to identify how disruptive or consolidating a publication or patent is within its respective field.
The package builds upon the original `node2vec` code by @skojaku, which implements the directional skip-gram algorithm. You can find the original `node2vec` implementation [here](https://github.com/skojaku/node2vec). `embedding-disruptiveness` modifies and extends this code to specifically calculate the disruption index and an embedding-based disruptiveness measure.
## Installation
To install the latest version of `embedding-disruptiveness`, run:
```bash
pip install --upgrade embedding-disruptiveness
```
## Requirements
This code requires at least two gpus
## Usage
Here is a basic example of how to use embedding-disruptiveness:
```python
import embedding_disruptiveness
# Initialize the model with required parameters
trainer = embedding_disruptiveness.EmbeddingTrainer(net_input = NETWORK_FILE_LOCATION , #(npz file type)
dim = 128, # dimension of embedding vectors
window_size=5, # windowsize
device_in = '6', # cuda device where in-vectors will be
device_out = '7',# cuda device where out-vectors will be
q_value = 1, # q value in the randomwalk
epochs =1, # epochsize
batch_size = 1024,# batchsize
save_dir = SAVE_LOCATION)
trainer.train()
```
## Model Parallelism
This package uses model parallelism to speed up the calculations, especially when dealing with large citation networks. Note: You need at least two GPUs to run the code. Ensure your environment is set up for multi-GPU usage, and the necessary CUDA drivers are installed.
Raw data
{
"_id": null,
"home_page": "https://github.com/yourusername/embedding-disruptiveness",
"name": "embedding-disruptiveness",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.8",
"maintainer_email": null,
"keywords": "science of science, disruption index, citation, innovation, edm, embedding disruptiveness measure",
"author": "Munjung Kim, Skojaku",
"author_email": "Munjung Kim <munjkim@iu.edu>, Skojaku <skojaku@binghamton.edu>",
"download_url": "https://files.pythonhosted.org/packages/da/4a/0407dcca141857da67b8a4f9589fe2a6771020b88423ab2b8d5167c276e3/embedding_disruptiveness-1.0.0.tar.gz",
"platform": null,
"description": "# embedding-disruptiveness\n\n`embedding-disruptiveness` is a Python package for calculating the Embedding Disruptiveness Index of papers or patents using a citation network. This measure helps to identify how disruptive or consolidating a publication or patent is within its respective field.\n\nThe package builds upon the original `node2vec` code by @skojaku, which implements the directional skip-gram algorithm. You can find the original `node2vec` implementation [here](https://github.com/skojaku/node2vec). `embedding-disruptiveness` modifies and extends this code to specifically calculate the disruption index and an embedding-based disruptiveness measure.\n\n\n## Installation\n\nTo install the latest version of `embedding-disruptiveness`, run:\n\n```bash\npip install --upgrade embedding-disruptiveness\n```\n\n\n## Requirements\nThis code requires at least two gpus\n\n## Usage\nHere is a basic example of how to use embedding-disruptiveness:\n\n```python\nimport embedding_disruptiveness\n\n# Initialize the model with required parameters\ntrainer = embedding_disruptiveness.EmbeddingTrainer(net_input = NETWORK_FILE_LOCATION , #(npz file type)\n dim = 128, # dimension of embedding vectors\n window_size=5, # windowsize\n device_in = '6', # cuda device where in-vectors will be \n device_out = '7',# cuda device where out-vectors will be \n q_value = 1, # q value in the randomwalk\n epochs =1, # epochsize\n batch_size = 1024,# batchsize\n save_dir = SAVE_LOCATION)\n\ntrainer.train()\n```\n\n## Model Parallelism\n\nThis package uses model parallelism to speed up the calculations, especially when dealing with large citation networks. Note: You need at least two GPUs to run the code. Ensure your environment is set up for multi-GPU usage, and the necessary CUDA drivers are installed.\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "A Python package to calculate the disruption index and embedding disruptiveness measure using a citation network.",
"version": "1.0.0",
"project_urls": {
"Documentation": "https://github.com/MunjungKim/embedding-disruptiveness#readme",
"Homepage": "https://github.com/MunjungKim/embedding-disruptiveness",
"Repository": "https://github.com/MunjungKim/embedding-disruptiveness"
},
"split_keywords": [
"science of science",
" disruption index",
" citation",
" innovation",
" edm",
" embedding disruptiveness measure"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "53c700da4c76c38562f73fee2732233adc7e3696a9934d12f82da648b896c157",
"md5": "27cd69480eccee06d95e4a9945e68aff",
"sha256": "9eeaf5b10cec49b7e9918f79435c2177579270f211a9e2cbfb4eb961e7585516"
},
"downloads": -1,
"filename": "embedding_disruptiveness-1.0.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "27cd69480eccee06d95e4a9945e68aff",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8",
"size": 19364,
"upload_time": "2024-11-26T06:18:20",
"upload_time_iso_8601": "2024-11-26T06:18:20.339843Z",
"url": "https://files.pythonhosted.org/packages/53/c7/00da4c76c38562f73fee2732233adc7e3696a9934d12f82da648b896c157/embedding_disruptiveness-1.0.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "da4a0407dcca141857da67b8a4f9589fe2a6771020b88423ab2b8d5167c276e3",
"md5": "896d564fe0d322551ec4f2b81610d340",
"sha256": "e85c2d60ecb1133442b888b1bb4bbf42f97a8fb0859617dc0854925aee557526"
},
"downloads": -1,
"filename": "embedding_disruptiveness-1.0.0.tar.gz",
"has_sig": false,
"md5_digest": "896d564fe0d322551ec4f2b81610d340",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8",
"size": 16328,
"upload_time": "2024-11-26T06:18:21",
"upload_time_iso_8601": "2024-11-26T06:18:21.821469Z",
"url": "https://files.pythonhosted.org/packages/da/4a/0407dcca141857da67b8a4f9589fe2a6771020b88423ab2b8d5167c276e3/embedding_disruptiveness-1.0.0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-11-26 06:18:21",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "yourusername",
"github_project": "embedding-disruptiveness",
"github_not_found": true,
"lcname": "embedding-disruptiveness"
}
JSON
