Can Artificial Intelligence (AI) Translate Arabic Abu-Brand Names with Different Prompts

In addition to “father”, أبو “Abu" in Arabic has several meanings, and is used in many contexts, including brand names. Despite their widespread use, there is a lack of studies that focus on the translation of Abu-brand names by Artificial Intelligence (AI). Therefore, this study investigated the translation of Abu-brand names by Microsoft Copilot (MC) and DeepSeek (DS) in terms of accuracy, translation strategies, causes of faulty equivalents and whether stakeholders can depend on AI in translating Abu-brand names. A sample of 100 Abu-brand names was collected and translated by MC and DS using three different tasks. In the three tasks, MC gave literal word-for-word translations of all 100 brand names, where it translated Abu as “father of” and the following noun semantically, whether the prompt specified the phrase type or not and whether each brand name was associated with the product name or not. Like MC, DS gave a literal word-for-word translation of all the brand names in tasks 1 and 2. However, when the product name was added, DS treated the brand names as Proper Nouns and transliterated them all in English regardless of whether they were coined by the manufacturers, business owners or consumers. However, 66% of the transliterations by DS were correctly used, but 34% were not supposed to be transliterated as they were grassroots (folk-coined) brand names, based on the packaging image, not the original English brand name which some consumers find difficult to pronounce and remember as a result, they coin a nickname based on the image on the packaging. Instead, the original English brand name should have been given by AI as an equivalent in the case of grassroots brand-names (Tiger balm instead of Abu Nimr Ointment for ابو نمر مرهم). Other interesting findings were that MC and DS gave identical English equivalents to 83% of the items in tasks 1 and 2. Additionally, DS gave double equivalents to 14.5% of the items in set 1, and faulty annotations/explanations of the brand name that show extraneous inferences and faulty guesses based on kunyas and nicknames that do not match a commercial context. Recommendations for stakeholders and for improving the performance of AI in translating Arabic Abu-brand names are given.

Authors

Institutions

Publication Details

Journal
Journal of Computer Science and Technology Studies
Published
2025-09-24
DOI
https://doi.org/10.32996/jcsts.2025.4.1.91
Primary Topic
Middle East and Rwanda Conflicts
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Can Artificial Intelligence (AI) Translate Arabic Abu-Brand Names with Different Prompts

Reima Al-Jarf
Journal of Computer Science and Technology Studies
Middle East and Rwanda Conflicts
article

Can Artificial Intelligence (AI) Translate Arabic Abu-Brand Names with Different Prompts

Reima Al-Jarf
article en

Abstract

In addition to “father”, أبو “Abu" in Arabic has several meanings, and is used in many contexts, including brand names. Despite their widespread use, there is a lack of studies that focus on the translation of Abu-brand names by Artificial Intelligence (AI). Therefore, this study investigated the translation of Abu-brand names by Microsoft Copilot (MC) and DeepSeek (DS) in terms of accuracy, translation strategies, causes of faulty equivalents and whether stakeholders can depend on AI in translating Abu-brand names. A sample of 100 Abu-brand names was collected and translated by MC and DS using three different tasks. In the three tasks, MC gave literal word-for-word translations of all 100 brand names, where it translated Abu as “father of” and the following noun semantically, whether the prompt specified the phrase type or not and whether each brand name was associated with the product name or not. Like MC, DS gave a literal word-for-word translation of all the brand names in tasks 1 and 2. However, when the product name was added, DS treated the brand names as Proper Nouns and transliterated them all in English regardless of whether they were coined by the manufacturers, business owners or consumers. However, 66% of the transliterations by DS were correctly used, but 34% were not supposed to be transliterated as they were grassroots (folk-coined) brand names, based on the packaging image, not the original English brand name which some consumers find difficult to pronounce and remember as a result, they coin a nickname based on the image on the packaging. Instead, the original English brand name should have been given by AI as an equivalent in the case of grassroots brand-names (Tiger balm instead of Abu Nimr Ointment for ابو نمر مرهم). Other interesting findings were that MC and DS gave identical English equivalents to 83% of the items in tasks 1 and 2. Additionally, DS gave double equivalents to 14.5% of the items in set 1, and faulty annotations/explanations of the brand name that show extraneous inferences and faulty guesses based on kunyas and nicknames that do not match a commercial context. Recommendations for stakeholders and for improving the performance of AI in translating Arabic Abu-brand names are given.

Journal of Computer Science and Technology StudiesVol. 7(9)
King Saud University (SA)
Openalex Percentile: Top 65%
Middle East and Rwanda Conflicts
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.