Sentiment and Emotion Classification of Indonesian E-Commerce Reviews via Multi-Task BiLSTM and AutoML Benchmarking
Hermawan Manurung, Ahmad Rizqi, Ibrahim Al-Kahfi, Martin Clinton Tosima Manullang, Ph.D · 2026
Read Sentiment and Emotion Classification of Indonesian E-Commerce Reviews via Multi-Task BiLSTM and AutoML Benchmarking by Hermawan Manurung, Ahmad Rizqi, Ibrahim Al-Kahfi, Martin Clinton Tosima Manullang, Ph.D (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-24720.txt.
Description
Manurung, Al-Kahfi, Rizqi and Manullang tackle sentiment and emotion classification of Indonesian e-commerce reviews, where slang, regional loanwords, numeric shorthand and emoji defeat lexicon tools. Their two-track pipeline uses the PRDECT-ID dataset. It holds 5,400 reviews across 29 categories, labeled for sentiment and five emotions. One track combines TF-IDF with an AutoML sweep. The other uses a bidirectional LSTM with two output heads. A preprocessing module applies 14 cleaning steps, among them a slang dictionary of 140 entries. Both tracks deploy as Gradio apps.
Catalog record
| CREATOR | Hermawan Manurung, Ahmad Rizqi, Ibrahim Al-Kahfi, Martin Clinton Tosima Manullang, Ph.D |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 20122 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.24720