Facebook Translation Service (FTS) Usage among Jordanians during COVID-19 Lockdown

Abstract

The world is facing an unprecedented virus outbreak, COVID-19, hitting more than 200 countries.Governments have been striving to prevent the spread of the virus through the lockdown.During the strict lockdown in Jordan, people needed to stay home, and they used available social networks to keep updated on COVID-19, with Facebook, the most popular social media platform.The study aimed to elicit information about assessing the use of FTS as a source of information in general and on COVID-19, in particular, FTS for those interested in English posts.However, it cannot read them and how reliable users think FTS is.The questionnaire was sent through the available networking sites, such as Facebook Messenger.The study found that 94.3% use Facebook daily; 87.1% of the participants activated Facebook Translation Service (FTS).It is found that 62.2% of the participants considered Facebook as a primary source of information regarding COVID-19 and 27.8% as secondary source.In terms of FTS usage, 87.3% used FTS in translating English Facebook posts into Arabic, and 83.8% used FTS in translating English Facebook COVID-19 posts into Arabic during the lockdown.On the other hand, it is found that the majority found FTS committed minor errors in terms of adequacy and fluency.This success is due to the usage of Neural Machine Translation (NMT)approach and bilingual text corpora.The advantage is that FTS is a well-trained database that can provide more accurate translation than other model.In conclusion, disregarding FTS output quality, our research shows that Facebook and FTS became a significant source of information during abrupt crises.Such research would encourage government officials to better use Facebook and FTS as complements to their national health campaigns.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 22 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.