text
stringlengths
24
32
https://openalex.org/W1976327432
https://openalex.org/W4296805512
https://openalex.org/W2069342313
https://openalex.org/W2494658236
https://openalex.org/W2080020695
https://openalex.org/W2038202938
https://openalex.org/W2043009264
https://openalex.org/W1972072419
https://openalex.org/W2086006935
https://openalex.org/W2590055506
https://openalex.org/W2166347821
https://openalex.org/W2092628464
https://openalex.org/W4319337429
https://openalex.org/W2412439850
https://openalex.org/W2024344593
https://openalex.org/W2347090135
https://openalex.org/W2082303365
https://openalex.org/W1968647885
https://openalex.org/W2322064794
https://openalex.org/W4287590154
https://openalex.org/W3106795082
https://openalex.org/W4292959150
https://openalex.org/W4221147640
https://openalex.org/W4321748483
https://openalex.org/W4295125129
https://openalex.org/W2048841252
https://openalex.org/W1981189573
https://openalex.org/W2154715074
https://openalex.org/W1993776840
https://openalex.org/W1985008026
https://openalex.org/W1998073894
https://openalex.org/W1997474467
https://openalex.org/W2133470139
https://openalex.org/W2099912508
https://openalex.org/W2059218921
https://openalex.org/W2810279113
https://openalex.org/W2729050566
https://openalex.org/W2312225717
https://openalex.org/W3124880449
https://openalex.org/W2161084662
https://openalex.org/W1972272448
https://openalex.org/W2763065452
https://openalex.org/W2009829463
https://openalex.org/W2071378688
https://openalex.org/W1982992065
https://openalex.org/W2070247214
https://openalex.org/W2318599676
https://openalex.org/W2017583196
https://openalex.org/W3111562161
https://openalex.org/W2139749834
https://openalex.org/W2082284474
https://openalex.org/W4316116532
https://openalex.org/W4327988895
https://openalex.org/W2028428499
https://openalex.org/W4206672451
https://openalex.org/W41495955
https://openalex.org/W4319812258
https://openalex.org/W2561303607
https://openalex.org/W2085284509
https://openalex.org/W3100866449
https://openalex.org/W2115739690
https://openalex.org/W2221336447
https://openalex.org/W2147370992
https://openalex.org/W2084527448
https://openalex.org/W3174912777
https://openalex.org/W1968060194
https://openalex.org/W2885794246
https://openalex.org/W83335121
https://openalex.org/W1545863082
https://openalex.org/W2085844340
https://openalex.org/W2072582976
https://openalex.org/W2090527912
https://openalex.org/W4285397066
https://openalex.org/W2800001287
https://openalex.org/W2019966600
https://openalex.org/W4286512634
https://openalex.org/W2162118735
https://openalex.org/W2528447115
https://openalex.org/W1986686461
https://openalex.org/W4304689648
https://openalex.org/W2978542003
https://openalex.org/W1612361173
https://openalex.org/W1547721749
https://openalex.org/W3151155428
https://openalex.org/W1628597191
https://openalex.org/W2199439547
https://openalex.org/W1820546952
https://openalex.org/W2292983454
https://openalex.org/W2952528795
https://openalex.org/W1520470547
https://openalex.org/W2952791373
https://openalex.org/W1930405358
https://openalex.org/W1529114714
https://openalex.org/W1993840381
https://openalex.org/W2897920659
https://openalex.org/W1123877853
https://openalex.org/W2341665396
https://openalex.org/W2482940081
https://openalex.org/W2039897610
https://openalex.org/W2413994661

abstracts-embeddings

This is the embeddings of the titles and abstracts of 95 million academic publications taken from the OpenAlex dataset as of May 5, 2023. The script that generated the embeddings is available on Github, but the general process is as follows:

  1. Reconstruct the text of the abstract from the inverted index format
  2. Construct a single document string in the format title + ' ' + abstract or just abstract if there is no title
  3. Determine if the document string is in English using fastText
  4. If it is in English, compute an embedding using the all-MiniLM-L6-v2 model provided by sentence-transformers

Though the OpenAlex dataset records 240 million works, not all of these works have abstracts or are in English. However, the all-MiniLM-L6-v2 model was only trained on English texts, hence the filtering.

Dataset Structure

In the future, this dataset might become a parquet in order to admit all the features offered by Hugging Face Datasets, but it consists only of a text file and a numpy memmap for now. The memmap is an array of many length-384 np.float16 vectors, and the i-th row vector in this array corresponds with the i-th line in the text file. The text file is just a list of ids that can be used to get more information from the OpenAlex API.

import numpy as np

with open('openalex_ids.txt', 'r') as f:
    idxs = f.read().splitlines()

embeddings = np.memmap('embeddings.memmap', dtype=np.float16, mode='r').reshape(-1, 384)

However, the memmap cannot be uploaded to Hugging Face as a single file, so it's split with the command split -b 3221225472 -d --suffix-length=3 --additional-suffix=.memmap embeddings.memmap embeddings_. It can be put back together with the command cat embeddings_*.memmap > embeddings.memmap.

Downloads last month
0
Edit dataset card