tags:
- codemixed
- dataset
- roman
- bengali
- english
- transaltion
- transliteration
- bengali pattern
- codemixed english
- pattern
RBE_Sent Dataset Description:
The RBE_Sent (Roman Bengali-English Sentiment) dataset is a synthetic, gold-standard code-mixed dataset developed for sentiment analysis tasks involving Romanized Bengali and English. It captures real-world bilingual usage by blending Roman Bengali with English tokens within the same textual instances. The dataset is designed to support research in multilingual natural language processing, especially in the context of low-resource, code-mixed languages. Each entry in RBE_Sent is annotated for sentiment, enabling supervised learning and evaluation. By providing a structured and labeled resource in this underexplored linguistic domain, RBE_Sent contributes to advancing computational methods for understanding Bengali-English code-mixed communication.
This Dataset contains Bengali matrix codemixed product review text. Dataset has 18086 roman Bengali-English codemixed product reviews .
license: mit language:
- bn
- en tags:
- codemixed
- product
- review
- ecommerce
- bengali
- dataset pretty_name: