Word Embedding Techniques for Classification of Star Ratings
Hesham Abdelmotaleb [email protected], Craig McNeile [email protected], Małgorzata Wojtyś [email protected] · 2025
Read Word Embedding Techniques for Classification of Star Ratings by Hesham Abdelmotaleb [email protected], Craig McNeile [email protected], Małgorzata Wojtyś [email protected] (2025) 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-2504-13653.txt.
Description
Phone firms learn a lot from the reviews their customers post online, if the star ratings in those reviews can be sorted by machine. Hesham Abdelmotaleb, Craig McNeile, and Malgorzata Wojtys test how the choice of word embedding shapes that sorting in a 2025 paper. Word embeddings turn text into number lists that capture word meaning.
The team built a fresh dataset of telecom customer reviews and tried several embedding methods, among them BERT, Word2Vec, and Doc2Vec. Each method was paired with a range of classifiers, and the team also tested ways to shrink the data with principal component methods. They tracked accuracy scores and the energy each method used. BERT paired with their component method gave the best scores on the harder tasks, while other pairings did better on simpler ones.
Catalog record
| CREATOR | Hesham Abdelmotaleb [email protected], Craig McNeile [email protected], Małgorzata Wojtyś [email protected] |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 70 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-13653