HYBRID CNN AND RNNS MODEL FOR SENTIMENT ANALYSIS

SITANGGANG, ANDREAS PERMANA PUTRA (2023) HYBRID CNN AND RNNS MODEL FOR SENTIMENT ANALYSIS. Skripsi thesis, UNIVERSITAS KATOLIK SOEGIJAPRANATA.

[img]
Preview
 Text
19.K1.0063-ANDREAS PERMANA PUTRA SITANGGANG_COVER_1.pdf
Download (196kB) | Preview
[img] Text
19.K1.0063-ANDREAS PERMANA PUTRA SITANGGANG_BAB I_1.pdf
Restricted to Registered users only
Download (170kB)
[img] Text
19.K1.0063-ANDREAS PERMANA PUTRA SITANGGANG_BAB II_1.pdf
Restricted to Registered users only
Download (174kB)
[img] Text
19.K1.0063-ANDREAS PERMANA PUTRA SITANGGANG_BAB III_1.pdf
Restricted to Registered users only
Download (589kB)
[img] Text
19.K1.0063-ANDREAS PERMANA PUTRA SITANGGANG_BAB IV_1.pdf
Restricted to Registered users only
Download (325kB)
[img] Text
19.K1.0063-ANDREAS PERMANA PUTRA SITANGGANG_BAB V_1.pdf
Restricted to Registered users only
Download (229kB)
[img]
Preview
 Text
19.K1.0063-ANDREAS PERMANA PUTRA SITANGGANG_DAPUS_1.pdf
Download (233kB) | Preview
[img] Text
19.K1.0063-ANDREAS PERMANA PUTRA SITANGGANG_LAMPIRAN_1.pdf
Restricted to Registered users only
Download (546kB)

Abstract

In the realm of sentiment analysis, understanding public opinion and customer feedback is of paramount importance. This study researches into the performance of a hybrid model that fuses Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) for sentiment analysis, considering various datasets. The findings consistently highlight the remarkable enhancement in sentiment classification accuracy achieved by the hybrid model in comparison to basic models, owing to its ability to capture complex data patterns. While the hybrid models tend to require slightly more training time, the trade-off between accuracy and training time remains manageable. Furthermore, the hybrid models outperform basic models across datasets.

Item Type: Thesis (Skripsi)
Subjects: 000 Computer Science, Information and General Works
Divisions: Faculty of Computer Science > Department of Informatics Engineering
Depositing User: Mr Yosua Norman Rumondor
Date Deposited: 05 Apr 2024 00:54
Last Modified: 05 Apr 2024 00:54
URI: http://repository.unika.ac.id/id/eprint/35177

Actions (login required)

View Item View Item
Unika Repository is powered by EPrints 3 which is developed by the School of Electronics and Computer Science at the University of Southampton. More information and software credits.