Drawing on both Israeli news media and Israeli social media sources, FilterLabs has been analyzing sentiment in articles and online discussion about the IDF.
The Israeli Defence Forces (IDF) are a pillar of Israeli life. Most Israelis serve in the IDF, or at least know others who do, so it should be no surprise that the institution enjoys a high degree of trust. At the same time, although the IDF’s approval level continues to be high, it has been dropping in recent years.
How will the war with Hamas affect this trend? The IDF’s leadership is taking the blame for military failures during the October 7th attacks, but whether or not the war will change the Israeli public’s perception of the IDF itself is a more difficult question.
FilterLabs.AI is in a position to shine some light on the subject. Over the past few months, we have been developing Arabic- and Hebrew-language AI tools for Talisman, our data analytics platform. Talisman can now track sentiment in a growing body of Israeli and Arabic-language media. So this week, drawing on both Israeli news media and Israeli social media sources, FilterLabs has been analyzing sentiment in articles and online discussion about the IDF.
Here is a graph of sentiment in mainstream news sources in Israel about the IDF:
The graph shows some dramatic swings. Sentiment rises quickly and stays high through mid-March, before taking a dive later in the month and rebounding again in early April.
Looking more closely at some of the underlying artifacts, FilterLabs analysts recognized a clear pattern. Military success reliably drove sentiment scores upwards. Positive news stories included IDF attacks on Hezbollah in Lebanon, strikes against targets in Gaza, and operations against Syrian army sites.
However, it would be a mistake to think that dropping sentiment corresponded with criticism of the military. Quite the opposite. While sentiment trended negative from March 20 to April 3, when we took a closer look at the data we found ongoing positive attitudes toward the IDF itself. Articles discussing withdrawals or pauses tended to have lower sentiment than those reporting increasing attacks. Some of the negative sentiment coincided with IDF troops withdrawing from Al-Shifa Hospital in Gaza, on April 1. Other pieces scoring negatively mentioned negative aspects of the overall situation, such as coverage of a community celebration of Purim “under the shadow of war.” But our spot checking consistently turned up supportive and positive discussion of the IDF. For example, while coverage of the Al-Shifa withdrawal registered lower sentiment than coverage of military advances, the event was not reported as a failure. Articles characterize the Al-Shifa operation as a success and mention several successful hits on "terrorist targets."
In sum, what was good news for the IDF correlated with positive sentiment in the news coverage. And even when there were dips in sentiment, a closer look at the artifacts Talisman unearthed revealed few if any cracks in the Israeli support of IDF operations as reflected in its news media.
The picture on social media was a little more complicated:
As in mainstream sources, there was a sentiment dip in late March and a quick rise in early April. There were some similar stories, too. The early April rise, for example, corresponded to stories about IDF operations like attacks into Lebanon.
The more negative stories, however, were different. Individual Israelis’ conversations online were more likely to express concern about whether or not Netanyahu would be reelected; in posts with negative emotion that mentioned both the IDF and Netanyahu, typically it was the prime minister and not the military that was the object of criticism.
Talisman’s data and analytical capabilities enabled us to dig into sentiment about the IDF in both mainstream news media and social media sources and examine differences between the two. While we found some similarities, we also identified an important point of divergence: online chatter was more likely to take up the politics of the war and criticize IDF leadership. By allowing users to look at the underlying data, Talisman enables more granular and verifiable analysis on topics and issues that are complicated and easily misinterpreted when we look at sentiment scores alone.
Have you taken the FilterLabs Newsletter survey? If not, please share your thoughts using the two-minute poll linked below. You'll help inform the focus of future newsletters—and as a gesture of our thanks, we'll set you up with a free one-month subscription to Talisman so you can try it out for yourself!
Fine print: Free Talisman trial is available to FilterLabs Newsletter subscribers only. (Not a subscriber? Sign up now—it's free!) Responses must be received by May 15, 2024. Valid survey submission and email address required.