refy
A scientific papers recommendation tool.
Overview
refy leverages Natural Langual Processing (NLP) machine learning tools to find new papers that might be relevant given the ones that you've read already.
There's a few software tools out there that facilitate the exploration of scientific literature, including:
- meta.org which allows users to set up feeds that identify newly published papers that seem relevant given a set of keywords
- inciteful and scite.ai let you explore the network of citations around a given paper of interest
- connected papers let's you visualize a graph representation of papers related to a given paper of interest
Most currently available software is limited in two key ways:
- Tools like meta.org rely on keywords, but keywords (e.g. computational neuroscience, Parkinson's Disease) are often overly general. As a result of that you have to sift through a lot irrelevant literature before you find something interesting
- Other tools like connected papers only work with one input paper at the time: you give it the title of a paper you've read and they give you suggestions. This is limiting: any software that can analyse all papers you've read can use a lot more information to find new papers that match more closely your interests.
This is what refy is for: refy analyzes the abstracts of several papers of yours and matches them agaist published preprints. By using many input papers at once refy has a lot more information at its disposal which (hopefully) means that it can better recommend relevant papers. By using the abstracts and not the paper titles, authors or keywords, refy focuses exclusively on the content of an article and has access to a wealth of data.
Refy downloads recently published preprints from BiorXiv and ArXiv, we thank BiorXiv and ArXiv for the API services they made freely available.
Usage
Installation
If you have an environment with python >= 3.6, you can install refy with:
pip install refy
getting suggested papers
import refy
d = refy.Recomender(
'library.bib', # path to your .bib file
n_days=30, # fetch preprints from the last N days
html_path="test.html", # save results to a .html (Optional)
N=10 # number of recomended papers
)
1 maintainer
Metadata
1 maintainer
Release files for refy 1.0.0.8
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| refy-1.0.0.8.tar.gz | 66.2 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| refy-1.0.0.8-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 83.1 kB
Release files / refy-1.0.0.8.tar.gz
| Download URL | refy-1.0.0.8.tar.gz |
|---|---|
| Size | 66.2 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
02d5c1d4298a9488d6b02c07c30e717ea5113b25e2902654f9352d6c3c197674
|
|
BLAKE2b-256 checksum How to use checksums |
98947a80163bf5c37b05f4febdd544a47c358fb27dc74be92752ff6f6b5774e0
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/3.1.1 pkginfo/1.5.0.1 requests/2.22.0 setuptools/53.0.0 requests-toolbelt/0.9.1 tqdm/4.54.0 CPython/3.7.4
|
Release files / refy-1.0.0.8-py3-none-any.whl
| Download URL | refy-1.0.0.8-py3-none-any.whl |
|---|---|
| Size | 16.9 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
71cefc306732ef1e04c9bbd148fcfa027d48a5a528174b3984d1afa9e470a002
|
|
BLAKE2b-256 checksum How to use checksums |
c8ada39d8603e2f672ccd088780b518569b3089bdd5d24b039f1f9051d666435
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/3.1.1 pkginfo/1.5.0.1 requests/2.22.0 setuptools/53.0.0 requests-toolbelt/0.9.1 tqdm/4.54.0 CPython/3.7.4
|
