Co-occurrence Weight Selection for Word Embeddings to Enhance Test Performance
This study revisits the problem of maximizing the performance of mathematical word representations for a given task. It is aimed to improve performance in analogy and similarity tasks by suggesting innovative weights instead of the counting weights used conventionally in counting-based methods of ge...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
Bursa Uludag University
2018-04-01
|
Series: | Uludağ University Journal of The Faculty of Engineering |
Subjects: | |
Online Access: | https://dergipark.org.tr/tr/pub/uumfd/issue/36268/318615 |
_version_ | 1797920056354537472 |
---|---|
author | Aykut Koç Veysel Yücesoy |
author_facet | Aykut Koç Veysel Yücesoy |
author_sort | Aykut Koç |
collection | DOAJ |
description | This study revisits
the problem of maximizing the performance of mathematical word representations
for a given task. It is aimed to improve performance in analogy and similarity
tasks by suggesting innovative weights instead of the counting weights used
conventionally in counting-based methods of generating word representations
(adding the statistics of word co-occurrences to the account). The language of
study was selected as Turkish. The root structures of Turkish words were managed
during the compilation of corpus such that each word having a suffix was
considered as a new word. The performance of the proposed co-occurrence weights
are analyzed with respect to the varying parameter and the results are
presented within the paper. |
first_indexed | 2024-04-10T13:55:06Z |
format | Article |
id | doaj.art-3e763673aed2404fb475327e93878492 |
institution | Directory Open Access Journal |
issn | 2148-4147 2148-4155 |
language | English |
last_indexed | 2024-04-10T13:55:06Z |
publishDate | 2018-04-01 |
publisher | Bursa Uludag University |
record_format | Article |
series | Uludağ University Journal of The Faculty of Engineering |
spelling | doaj.art-3e763673aed2404fb475327e938784922023-02-15T16:10:29ZengBursa Uludag UniversityUludağ University Journal of The Faculty of Engineering2148-41472148-41552018-04-01231314010.17482/uumfd.3186151779Co-occurrence Weight Selection for Word Embeddings to Enhance Test PerformanceAykut Koç0Veysel Yücesoy1ASELSANASELSANThis study revisits the problem of maximizing the performance of mathematical word representations for a given task. It is aimed to improve performance in analogy and similarity tasks by suggesting innovative weights instead of the counting weights used conventionally in counting-based methods of generating word representations (adding the statistics of word co-occurrences to the account). The language of study was selected as Turkish. The root structures of Turkish words were managed during the compilation of corpus such that each word having a suffix was considered as a new word. The performance of the proposed co-occurrence weights are analyzed with respect to the varying parameter and the results are presented within the paper.https://dergipark.org.tr/tr/pub/uumfd/issue/36268/318615kelime temsilleridoğal dil işlemei̇statistiksel dilbilimiword embeddingsnatural language processingstatistical linguistics |
spellingShingle | Aykut Koç Veysel Yücesoy Co-occurrence Weight Selection for Word Embeddings to Enhance Test Performance Uludağ University Journal of The Faculty of Engineering kelime temsilleri doğal dil işleme i̇statistiksel dilbilimi word embeddings natural language processing statistical linguistics |
title | Co-occurrence Weight Selection for Word Embeddings to Enhance Test Performance |
title_full | Co-occurrence Weight Selection for Word Embeddings to Enhance Test Performance |
title_fullStr | Co-occurrence Weight Selection for Word Embeddings to Enhance Test Performance |
title_full_unstemmed | Co-occurrence Weight Selection for Word Embeddings to Enhance Test Performance |
title_short | Co-occurrence Weight Selection for Word Embeddings to Enhance Test Performance |
title_sort | co occurrence weight selection for word embeddings to enhance test performance |
topic | kelime temsilleri doğal dil işleme i̇statistiksel dilbilimi word embeddings natural language processing statistical linguistics |
url | https://dergipark.org.tr/tr/pub/uumfd/issue/36268/318615 |
work_keys_str_mv | AT aykutkoc cooccurrenceweightselectionforwordembeddingstoenhancetestperformance AT veyselyucesoy cooccurrenceweightselectionforwordembeddingstoenhancetestperformance |