Ensemble Deep Learning for Multilabel Binary Classification of User-Generated Content
Sentiment analysis usually refers to the analysis of human-generated content via a polarity filter. Affective computing deals with the exact emotions conveyed through information. Emotional information most frequently cannot be accurately described by a single emotion class. Multilabel classifiers c...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-04-01
|
| Series: | Algorithms |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1999-4893/13/4/83 |
| _version_ | 1797571775228280832 |
|---|---|
| author | Giannis Haralabopoulos Ioannis Anagnostopoulos Derek McAuley |
| author_facet | Giannis Haralabopoulos Ioannis Anagnostopoulos Derek McAuley |
| author_sort | Giannis Haralabopoulos |
| collection | DOAJ |
| description | Sentiment analysis usually refers to the analysis of human-generated content via a polarity filter. Affective computing deals with the exact emotions conveyed through information. Emotional information most frequently cannot be accurately described by a single emotion class. Multilabel classifiers can categorize human-generated content in multiple emotional classes. Ensemble learning can improve the statistical, computational and representation aspects of such classifiers. We present a baseline stacked ensemble and propose a weighted ensemble. Our proposed weighted ensemble can use multiple classifiers to improve classification results without hyperparameter tuning or data overfitting. We evaluate our ensemble models with two datasets. The first dataset is from Semeval2018-Task 1 and contains almost 7000 Tweets, labeled with 11 sentiment classes. The second dataset is the Toxic Comment Dataset with more than 150,000 comments, labeled with six different levels of abuse or harassment. Our results suggest that ensemble learning improves classification results by <inline-formula> <math display="inline"> <semantics> <mrow> <mn>1.5</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula> to <inline-formula> <math display="inline"> <semantics> <mrow> <mn>5.4</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula>. |
| first_indexed | 2024-03-10T20:46:20Z |
| format | Article |
| id | doaj.art-da6b6f6054b94f17a7d16afb17e83577 |
| institution | Directory Open Access Journal |
| issn | 1999-4893 |
| language | English |
| last_indexed | 2024-03-10T20:46:20Z |
| publishDate | 2020-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Algorithms |
| spelling | doaj.art-da6b6f6054b94f17a7d16afb17e835772023-11-19T20:21:10ZengMDPI AGAlgorithms1999-48932020-04-011348310.3390/a13040083Ensemble Deep Learning for Multilabel Binary Classification of User-Generated ContentGiannis Haralabopoulos0Ioannis Anagnostopoulos1Derek McAuley2School of Computer Science, University of Nottingham, Nottingham NG8 1BB, UKDepartment of Computer Science and Biomedical Informatics, University of Thessaly, 35131 Lamia, GreeceSchool of Computer Science, University of Nottingham, Nottingham NG8 1BB, UKSentiment analysis usually refers to the analysis of human-generated content via a polarity filter. Affective computing deals with the exact emotions conveyed through information. Emotional information most frequently cannot be accurately described by a single emotion class. Multilabel classifiers can categorize human-generated content in multiple emotional classes. Ensemble learning can improve the statistical, computational and representation aspects of such classifiers. We present a baseline stacked ensemble and propose a weighted ensemble. Our proposed weighted ensemble can use multiple classifiers to improve classification results without hyperparameter tuning or data overfitting. We evaluate our ensemble models with two datasets. The first dataset is from Semeval2018-Task 1 and contains almost 7000 Tweets, labeled with 11 sentiment classes. The second dataset is the Toxic Comment Dataset with more than 150,000 comments, labeled with six different levels of abuse or harassment. Our results suggest that ensemble learning improves classification results by <inline-formula> <math display="inline"> <semantics> <mrow> <mn>1.5</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula> to <inline-formula> <math display="inline"> <semantics> <mrow> <mn>5.4</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula>.https://www.mdpi.com/1999-4893/13/4/83ensemble learningsentiment analysismultilabel classificationdeep neural networkspure emotionSemeval 2018 Task 1 |
| spellingShingle | Giannis Haralabopoulos Ioannis Anagnostopoulos Derek McAuley Ensemble Deep Learning for Multilabel Binary Classification of User-Generated Content Algorithms ensemble learning sentiment analysis multilabel classification deep neural networks pure emotion Semeval 2018 Task 1 |
| title | Ensemble Deep Learning for Multilabel Binary Classification of User-Generated Content |
| title_full | Ensemble Deep Learning for Multilabel Binary Classification of User-Generated Content |
| title_fullStr | Ensemble Deep Learning for Multilabel Binary Classification of User-Generated Content |
| title_full_unstemmed | Ensemble Deep Learning for Multilabel Binary Classification of User-Generated Content |
| title_short | Ensemble Deep Learning for Multilabel Binary Classification of User-Generated Content |
| title_sort | ensemble deep learning for multilabel binary classification of user generated content |
| topic | ensemble learning sentiment analysis multilabel classification deep neural networks pure emotion Semeval 2018 Task 1 |
| url | https://www.mdpi.com/1999-4893/13/4/83 |
| work_keys_str_mv | AT giannisharalabopoulos ensembledeeplearningformultilabelbinaryclassificationofusergeneratedcontent AT ioannisanagnostopoulos ensembledeeplearningformultilabelbinaryclassificationofusergeneratedcontent AT derekmcauley ensembledeeplearningformultilabelbinaryclassificationofusergeneratedcontent |