Insights from FilterLabs’ analysis of Arabic-language news
In the current Israel-Palestine crisis, one of the most important questions is how the broader region will respond. Will Arab nations remain neutral? Condemn Israel and send humanitarian aid to Palestine? Or is there a chance that they could enter the conflict themselves?
FilterLabs’ AI tools enable the company to collect sentiment data for online discourse in Arabic, among other languages. Recently, this has allowed us to follow shifts in sentiment in regional news in the weeks since the crisis began and look for clues to how nations may respond. For example, our team has been analyzing Jordanian Arabic-language media published since October 7th, and our analysts noticed an interesting pattern in articles mentioning Jordan’s King Abdullah II.
In the days following the October 7th attacks, sentiment scores in articles mentioning Abdullah II fell sharply. Then, in late October, sentiment suddenly shot up again.
It is important to note that negative sentiment scores in articles that mention Abdullah II does not necessarily mean the articles were critical of the king. Rather, it appears that the sentiment is reflecting how Abdullah II himself (and/or his communications team) has been talking about the Israel-Palestine crisis.
For several weeks after the October 7th attacks, the king emphasized the dangers of the war. Specifically, he warned it could escalate out of control and into a region-wide conflict. The articles were full of phrases like:
“...the situation will continue to worsen…”
“...fear of the spiral of violence…”
“...new cycle of violence…”
“...drag the region into catastrophe…”
The phrase “dangerous escalation,” in particular, popped up repeatedly. It seems that Abdullah II, at least as he was presented in the news stories, was missing no chance to warn other world leaders that the conflict could spiral beyond Israel-Palestine and draw in the entire region. FilterLabs.AI was picking up on the negativity in his warnings.
In late October, the sentiment in articles referring to Abdullah II turned in a positive direction. A sudden change in sentiment can suggest a concerted effort to change the media narrative, and that appears to be what was happening here.
New stories about Abdullah no longer included his warnings. Instead, they reported on what the king himself was doing about the crisis. They praised him for:
- Getting a UN resolution passed: “His Majesty King Abdullah II succeeded in issuing a resolution from the United Nations General Assembly calling for a ceasefire…”
- Sending humanitarian aid to Palestinians: “in response to directions from his majesty … 45 thousand tons of wheat and grains and 7 trucks of medicines, medical and pharmaceutical needs and equipment to our people in the West Bank…”
- Maintaining a sufficient stock of food/supplies so that Jordanians are not at risk in case of global economic disruption: “...the forward-looking vision of His Majesty King Abdullah II made Jordan immune to being affected by crisis.”
And so on.
Sentiment analysis cannot, on its own, tell observers what the leaders of the broader region are thinking. It can, however, pick up on narrative shifts. In this case, it appears that there has been a conscious choice on the part of the Jordanian monarchy to change how it talks about the war. Done carefully, sentiment analysis can catch these changes as they happen. Policymakers can receive early insight into foreign governments’ thinking, and formulate their own response.