Use of DA’s Sila to analyse Arabic dialects on social media

10th September 2020

Arabic, the second hardest language to learn,  with 30 modern dialects, remains one of the most complex languages in the world. Combined with the fact that 79% of the MENA internet users spend up to three hours a day updating their social platforms, we start to paint an interesting picture of a landscape that takes some understanding.

The use of Arabic dialects

While ‘everyday’ Arabic is used professionally in the media, dialects of the language are used by individuals on a daily basis, including on social media. The same applies to Arab influencers who prefer using dialects which they are more comfortable with.

Video by LearnArabicWithMaha showcasing the vast differences between Arabic dialects

Arabic dialects on social media

Here at Digital Ape, we are both passionate and curious about data and creators (influencers).

As most Arabic on social is in dialect, we created Sila – the <very first tool created for the Arabic-speaking world to provide audience-first social intelligence and creator campaign management to our clients.

Sila is updated daily with millions of posts and interactions. This way the data warehouse is always up to date with the most recent data using real-time reporting of Arabic content, which includes (but is not exclusive to) different Arabic dialects.

An example: Creator Sentiment

Using data science and natural language processing models based on Arabic dialects, we see audience’s reactions and their affinities. We also can understand a broader range of inputs to topics, making the landscape more complete.

Those models allow us to build on previous capabilities to see the audiences’ age, gender, location, and engagements and cross-compare. 

This enables analysis of the sentiment of the target audience. On a social network like Instagram, the slightest movement from negative to positive sentiment can mean a world of difference to ROI.

Influencer Sentiment D/A 2020

When coupled with real-time reporting, combined with natural-language-processing models, helps better refine and learn from the audiences’ reactions in an unbiased setting.

Takeaway

Sila analyses both dialects and standard Arabic as it updates millions of posts and comments on a daily basis. Having a tool at our disposal with a foundation built upon Arabic dialect prevents key influencer interactions from falling into the obscurity. It also allows us a more complete picture of a topic.

Through this, we produce sentiment reports which help us identify potential influencers and report on current collaborations. The enables us to measure ROI accurately based on campaign performance through the positivity of a campaign.

About the tool

D/A has built a custom data warehouse service called Sila that ingests thousands of social media data points a day from across the Arab world. Powered by data science we turn that raw social data into insights, strategy, and actionable marketing.

To find out more about how to best use an audience first approach to influencers, check out the Art & Science of Influencer Marketing in the G.C.C or get in touch.