[KongchangAI]
Product

ReciLex

A multilingual culinary lexicon containing approximately 15,000 translated entities across ten categories such as ingredients, utensils and chef actions, injected as external knowledge to enhance the cultural adaptation of recipe translation for low-resource languages.

Timeline (last 90 days)

Nov 18

The paper proposes ReciLex, a multilingual culinary lexicon containing about 15,000 translated entities spanning ten recipe entity categories.

Unverified50%
Nov 18

The role of ReciLex is to serve as injected external knowledge rather than to replace model training.

Unverified50%
Nov 18

The core question of the paper is whether lexicon augmentation alone, without large-scale collection of target-language recipe corpora and without fine-tuning, can improve the quality of culturally adapted translation.

Unverified50%
Nov 18

On the selected models and languages, ReciLex augmentation raised average BLEU from 13.57 to 14.51.

Unverified50%
Nov 18

On the selected models and languages, ReciLex augmentation raised chrF from 37.89 to 38.74.

Unverified50%
Nov 18

On the selected models and languages, ReciLex augmentation lowered TER from 77.14 to 75.12.

Unverified50%
Nov 18

The study focuses on three Nigerian languages and the specific domain of recipes, and whether the conclusions transfer to other low-resource languages or text types still requires subsequent verification.

Unverified50%

All Facts (7)

Source Articles