October 4, 2018

NEW: AI Event Detection in Arabic

Jordan Ching by Jordan Ching

Today, we’re excited to launch native Arabic language event detection in SAM! All SAM users will now be able to monitor Arabic-speaking regions with greater speed and accuracy than ever before – no translation skills required! Our Real-Time Knowledge Engine now reads tweets written in Arabic and breaks down each post to understand if a critical event has occurred and where it is happening.

Our Data and Engineering teams have worked hard to understand and map the unique aspects of the Arabic language to build our NLP and NER systems, as well as our processes for natively-written Arabic slang, geo-references, and colloquialisms. As a result, our users now have an extremely easy way to monitor critical events (from attacks to traffic incidents) that happen in Arabic-speaking regions before western media sources report on the events.

Arabic is a critical language for many businesses, newsrooms and analysts – it has been our most requested language to date. Yet many services or tools do not properly support Arabic monitoring despite it being the sixth-most-popular language on Twitter. We’re thrilled to offer this service to all our users, and deliver breaking news direct from eyewitnesses and local media sources.

Let’s take a look at how SAM stacks up and let the data speak for itself:

FIRE in Riyadh, Saudi Arabia: A huge fire broke out at a power plant in Riyadh sending a ball of flames 10 meters into the air. The blaze broke out inside a transformer at the plant in the city’s Al-Nafil district. The Saudi Electricity Company said no one had been hurt, but that power supplies to some nearby areas had been shut off. SAM alerted this event 36 minutes before local news outlet Al Riyadh and 2 hours and 2 minutes before national news outlet Arab News provided coverage.

CRASH at Khartoum International Airport, Khartoum, Sudan: A collision between two Sudanese army planes occurred on the runway at Khartoum airport, injuring eight people. The Soviet-era Antonov aircrafts landed within seconds of each other when the accident occurred. This accident is the third to hit Sudan’s aging military fleet in less than a month. SAM alerted to this event 28 minutes before news outlet Al Arabiya and 2 hours and 36 minutes before national news outlet Sudan News provided coverage.

In both of the cases highlighted above, western media sources, like BBC Arabic and Sky Arabia, did not cover the events. Because SAM taps into eye-witness accounts and hyperlocal news sources we’re able to provide consistent coverage without being reliant on the western media news agenda. With this update, SAM users not only heard about major events not covered elsewhere, but also got notified before the local press.

If you have a SAM account log in to see native Arabic language event detection in action!







By continuing to use this site you consent to the use of cookies in accordance with our cookie policy. Learn more