Twitter iStock

Over the course of 2016, a World Jewish Congress survey finds, more than 382,000 anti-Semitic posts were uploaded to Twitter, Facebook, YouTube, Instagram, blogs and other social media forums.

The research, conducted together with Vigo Social Intelligence, showed that an overwhelming 63 percent of all anti-Semitic content online can be found on Twitter.

More than 382,000 anti-Semitic posts were uploaded to social media in 2016, an average of one post every 83 seconds.

“We knew that anti-Semitism online was on the rise, but the numbers revealed in this report give us concrete data as to how alarming the situation really is,” World Jewish Congress CEO and Executive Vice President Robert R. Singer said. “We hope this serves as a wake-up call to all internet forums to maintain moral standards, rid themselves of offensive content, and make the digital world a safer place for all.”

The WJC research analyzed tens of millions of posts in 20 languages on various social media. After Twitter, the highest number of anti-Semitic posts online can be found on blogs, at 16 percent. 11 percent of anti-Semitic content online was posted to Facebook, followed by Instagram with 6 percent, YouTube with 2 percent, and 2 percent on other forums.

The criteria used to determine whether a post was anti-Semitic was based on International Holocaust Remembrance Alliance (IHRA)’s May 2016 definition. It states that “anti-Semitism is a certain perception of Jews, which may be expressed as hatred toward Jews. Rhetorical and physical manifestations of anti-Semitism are directed toward Jewish or non-Jewish individuals and/or their property, toward Jewish community institutions and religious facilities.”

Posts criticizing Israel or Israel’s activities were not counted in this study.

The anti-Semitic posts were identified through a database of searchable word phrases and terms online, then translated into leading languages and scanned through the internet. A representative sample was read and codified by analysts to refine searches and deepen qualitative analysis. A total of 7,600 posts were read in different languages, indicating a representative sample of 2% of the total discourse included in the analysis.