Predicting word vectors for microtext
The use of computer-mediated communication has resulted in a new form of written text called Microtext, which is very different from well-written text. Most previous approaches deal with microtext at the character level rather than just words resulting in increased processing time. In this paper, we...
Saved in:
Main Authors: | , , , |
---|---|
Other Authors: | |
Format: | Article |
Language: | English |
Published: |
2024
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/180129 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-180129 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-1801292024-09-20T15:35:57Z Predicting word vectors for microtext Chaturvedi, Iti Satapathy, Ranjan Lynch, Curtis Cambria, Erik School of Computer Science and Engineering Computer and Information Science Microtext Sentiment analysis The use of computer-mediated communication has resulted in a new form of written text called Microtext, which is very different from well-written text. Most previous approaches deal with microtext at the character level rather than just words resulting in increased processing time. In this paper, we propose to transform static word vectors to dynamic form by modelling the effect of neighbouring words and their sentiment strength in the AffectiveSpace. To evaluate the approach, we crawled Tweets from diverse topics and human annotation was used to label their sentiments. We also normalized the tweets to fix phonetic variations, spelling errors, and abbreviations manually. A total of 1432 out-of-vocabulary (OOV) texts and their IV texts made it to the final corpus with their corresponding polarity. To assess the quality of the corpus, we used several OOV classifiers such as linear regression and observed over 90% accuracy. Next, we inferred word vectors using a novel four-gram model based on sentiment intensity and reported accuracy on both open domain and closed domain sentiment classifiers. We observed an improvement in the range of 4–20 on Twitter, Movie and Airline reviews over baselines. Ministry of Education (MOE) Published version This research is partially supported by the Ministry of Education, Singapore under its MOE Academic Research Fund Tier 2 (STEM RIE2025 Award MOE‐T2EP20123‐0005). This work is also partially supported by the College of Science and Engineering at James Cook University. Open access publishing facilitated by James Cook University, as part of the Wiley - James Cook University agreement via the Council of Australian University Librarians. 2024-09-18T05:01:43Z 2024-09-18T05:01:43Z 2024 Journal Article Chaturvedi, I., Satapathy, R., Lynch, C. & Cambria, E. (2024). Predicting word vectors for microtext. Expert Systems, 41(8), 13589-. https://dx.doi.org/10.1111/exsy.13589 0266-4720 https://hdl.handle.net/10356/180129 10.1111/exsy.13589 2-s2.0-85189503180 8 41 13589 en MOE‐T2EP20123‐0005 Expert Systems © 2024 The Authors. Expert Systems published by John Wiley & Sons Ltd. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. application/pdf |
institution |
Nanyang Technological University |
building |
NTU Library |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
NTU Library |
collection |
DR-NTU |
language |
English |
topic |
Computer and Information Science Microtext Sentiment analysis |
spellingShingle |
Computer and Information Science Microtext Sentiment analysis Chaturvedi, Iti Satapathy, Ranjan Lynch, Curtis Cambria, Erik Predicting word vectors for microtext |
description |
The use of computer-mediated communication has resulted in a new form of written text called Microtext, which is very different from well-written text. Most previous approaches deal with microtext at the character level rather than just words resulting in increased processing time. In this paper, we propose to transform static word vectors to dynamic form by modelling the effect of neighbouring words and their sentiment strength in the AffectiveSpace. To evaluate the approach, we crawled Tweets from diverse topics and human annotation was used to label their sentiments. We also normalized the tweets to fix phonetic variations, spelling errors, and abbreviations manually. A total of 1432 out-of-vocabulary (OOV) texts and their IV texts made it to the final corpus with their corresponding polarity. To assess the quality of the corpus, we used several OOV classifiers such as linear regression and observed over 90% accuracy. Next, we inferred word vectors using a novel four-gram model based on sentiment intensity and reported accuracy on both open domain and closed domain sentiment classifiers. We observed an improvement in the range of 4–20 on Twitter, Movie and Airline reviews over baselines. |
author2 |
School of Computer Science and Engineering |
author_facet |
School of Computer Science and Engineering Chaturvedi, Iti Satapathy, Ranjan Lynch, Curtis Cambria, Erik |
format |
Article |
author |
Chaturvedi, Iti Satapathy, Ranjan Lynch, Curtis Cambria, Erik |
author_sort |
Chaturvedi, Iti |
title |
Predicting word vectors for microtext |
title_short |
Predicting word vectors for microtext |
title_full |
Predicting word vectors for microtext |
title_fullStr |
Predicting word vectors for microtext |
title_full_unstemmed |
Predicting word vectors for microtext |
title_sort |
predicting word vectors for microtext |
publishDate |
2024 |
url |
https://hdl.handle.net/10356/180129 |
_version_ |
1814047206653558784 |