Published November 6, 2019 | Version 1.0
Dataset Open

Swahili word analogy dataset

  • 1. University of Electronic Science and Technology of China

Description

Swahili Analogy dataset contains pairs of words that are organized in 4's to facilitate word analogy test. Word analogy test is used to evaluate the quality of word representation vectors from a language model. The dataset contains 12,864 questions that have been organized in 12 categories.

Files

swaanalogydata.txt

Files (425.4 kB)

Name Size Download all
md5:9147a5458450653b1cd4c52acf5c678a
425.4 kB Preview Download