Performance Evaluation of Word2Vec and FastText Embeddings in a CNN-BiLSTM Model for Sentiment Classification of the LPDP Alumni Controversy

Dwi Erzalianti; Joice Junansi Tandirerung; Cici Suhaeni; Bagus Sartono

doi:10.30871/jaic.v10i4.13298

Authors

Dwi Erzalianti IPB University
Joice Junansi Tandirerung IPB University
Cici Suhaeni IPB University
Bagus Sartono IPB University

DOI:

https://doi.org/10.30871/jaic.v10i4.13298

Keywords:

Sentiment Analysis, CNN-BiLSTM, Imbalanced Data, Word2Vec, FastText

Abstract

This study aims to analyze public sentiment toward the LPDP alumni controversy on social media using a deep learning approach. The research data consist of YouTube user comments related to the LPDP issue, which were processed through text preprocessing and automatically labeled using IndoBERT into three sentiment classes: negative, neutral, and positive. This study compares two text representation methods, namely Word2Vec and FastText, implemented within a hybrid CNN–BiLSTM architecture. In addition, data imbalance was addressed using class weighting and undersampling scenarios, while TF-IDF-based Logistic Regression was used as the baseline model. The results show that the baseline achieved an accuracy of 0.83 but was strongly biased toward the negative class as the majority class. The CNN–BiLSTM model improved the ability to detect minority classes. Under the class weighting scenario, FastText demonstrated more stable performance with an accuracy of 0.77 and a macro F1-score of 0.62. Under the undersampling scenario, Word2Vec was more stable, achieving an accuracy of 0.68 and a macro F1-score of 0.67. These findings indicate that both text representation and imbalance-handling strategies substantially affect sentiment classification performance.

Downloads

Download data is not yet available.

References

[1] D. A. Firdaus and E. B. Setiawan, “Analisis Sentimen pada X Terhadap Pilkada 2024 Menggunakan Ekspansi Fitur FastText dan CNN dengan Optimasi Bat Algorithm,” eProceedings of Engineering, 2025.

[2] A. R. Gunawan and R. F. Alfa Aziza, “Sentiment Analysis Using LSTM Algorithm Regarding Grab Application Services in Indonesia,” Journal of Applied Informatics and Computing, vol. 9, no. 2, pp. 322–332, Mar. 2025, doi: 10.30871/jaic.v9i2.8696.

[3] Handoko, Junadhi, Triyani Arita Fitri, and Agustin, “Optimization of Deep Learning with FastText for Sentiment Analysis of the SIREKAP 2024 Application,” The Indonesian Journal of Computer Science, vol. 14, no. 2, Apr. 2025, doi: 10.33022/ijcs.v14i2.4809.

[4] D. Ariyus, D. Manongga, and I. Sembiring, “Enhancing Sentiment Analysis of Indonesian Tourism Video Content Commentary on TikTok: A FastText and Bi-LSTM Approach,” Engineering, Technology & Applied Science Research, vol. 14, no. 6, pp. 18020–18028, Dec. 2024, doi: 10.48084/etasr.8859.

[5] M. E. Alzahrani, T. H. H. Aldhyani, S. N. Alsubari, M. M. Althobaiti, and A. Fahad, “Developing an Intelligent System with Deep Learning Algorithms for Sentiment Analysis of E-Commerce Product Reviews,” Comput. Intell. Neurosci., vol. 2022, pp. 1–10, May 2022, doi: 10.1155/2022/3840071.

[6] A. S. Mirdan, S. Buyrukoglu, and M. R. Baker, “Advanced deep learning techniques for sentiment analysis: combining Bi-LSTM, CNN, and attention layers,” International Journal of Advances in Intelligent Informatics, vol. 11, no. 1, p. 55, Feb. 2025, doi: 10.26555/ijain.v11i1.1848.

[7] L. Khan, A. Amjad, K. M. Afaq, and H.-T. Chang, “Deep Sentiment Analysis Using CNN-LSTM Architecture of English and Roman Urdu Text Shared in Social Media,” Applied Sciences, vol. 12, no. 5, p. 2694, Mar. 2022, doi: 10.3390/app12052694.

[8] V. R. Prasetyo, M. F. Naufal, and K. Wijaya, “Sentiment Analysis of ChatGPT on Indonesian Text using Hybrid CNN and Bi-LSTM,” Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi), vol. 9, no. 2, pp. 327–333, Apr. 2025, doi: 10.29207/resti.v9i2.6334.

[9] Merinda Lestandy and Abdurrahim, “Effect of Word2Vec Weighting with CNN-BiLSTM Model on Emotion Classification,” Jurnal Nasional Pendidikan Teknik Informatika (JANAPATI), vol. 12, no. 1, pp. 99–107, Mar. 2023, doi: 10.23887/janapati.v12i1.58571.

[10] L. N. Aqilla, Y. Sibaroni, and S. S. Prasetiyowati, “Word2vec Architecture in Sentiment Classification of Fuel Price Increase Using CNN-BiLSTM Method,” Sinkron, vol. 8, no. 3, pp. 1654–1664, Jul. 2023, doi: 10.33395/sinkron.v8i3.12639.

[11] Muhammad Taufik Maulana, Lailil Muflikhah, and Tirana Noor Fatyanosa, “Analisis Sentimen Pengguna Indodax Menggunakan FastText dan Convolutional Neural Network (CNN),” Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer (J-PTIIK), vol. 9, no. 6, Jun. 2025.

[12] A. S. Nugroho and K. Nugroho, “Comparison of RNN and LSTM Algorithms Based on Fasttext Embeddings in Sentiment Analysis on the Merdeka Mengajar Platform,” sinkron, vol. 9, no. 1, pp. 117–128, Jan. 2025, doi: 10.33395/sinkron.v9i1.14296.

[13] A. Saekhu, B. Berlilana, and D. I. S. Saputra, “Comparative Analysis of Data Balancing Techniques for Machine Learning Classification on Imbalanced Student Perception Datasets,” Jurnal Teknik Informatika (Jutif), vol. 6, no. 2, pp. 627–640, Apr. 2025, doi: 10.52436/1.jutif.2025.6.2.4286.

[14] C. Sintiya, G. H. Hutagaol, D. Bate`e, and S. Irviantina, “Evaluasi Teknik Resampling untuk Class Balancing dalam Analisis Sentimen Kesehatan Mental Berbasis Bi-LSTM,” Jurnal Sifo Mikroskil, vol. 26, no. 2, Oct. 2025, doi: 10.55601/jsm.v26i2.1799.

[15] Y. Yang, H. A. Khorshidi, and U. Aickelin, “A review on over-sampling techniques in classification of multi-class imbalanced datasets: insights for medical problems,” Front. Digit. Health, vol. 6, Jul. 2024, doi: 10.3389/fdgth.2024.1430245.

[16] M. Carvalho, A. J. Pinho, and S. Brás, “Resampling approaches to handle class imbalance: a review from a data perspective,” J. Big Data, vol. 12, no. 1, p. 71, Mar. 2025, doi: 10.1186/s40537-025-01119-4.