demodata

<p>This demo data contains a sample of 8000 English words
with 300-dimension word vectors pre-trained
using the "word2vec" algorithm based on the Google News corpus.
Most of these words are from the Top 8000 frequent wordlist,
whereas a few are selected from less frequent words and appended.</p>

An integrative toolbox of word embedding research that provides:
(1) a collection of 'pre-trained' static word vectors in the '.RData'
compressed format <https://psychbruce.github.io/WordVector_RData.pdf>;
(2) a series of functions to process, analyze, and visualize word vectors;
(3) a range of tests to examine conceptual associations, including
the Word Embedding Association Test <doi:10.1126/science.aal4230>
and the Relative Norm Distance <doi:10.1073/pnas.1720347115>,
with permutation test of significance;
(4) a set of training methods to locally train (static) word vectors
from text corpora, including 'Word2Vec' <arXiv:1301.3781>,
'GloVe' <doi:10.3115/v1/D14-1162>, and 'FastText' <arXiv:1607.04606>;
(5) a group of functions to download 'pre-trained' language models
(e.g., 'GPT', 'BERT') and extract contextualized (dynamic) word vectors
(based on the R package 'text').

Han-Wu-Shuang Bao

PsychWordVec

Word Embedding Research Framework for Psychological Science

demodata function

<p>A <code>data.table</code> (of new class <code>wordvec</code>) with two variables <code>word</code> and <code>vec</code>,
transformed from the raw data (see the URL in Source) into <code>.RData</code>
using the <code>data_transform</code> function.</p>

demodata: Demo data (pre-trained using word2vec on Google News; 8000 vocab, 300 dims).

Description

Usage

Arguments

Format

Examples