• by Laurie Sullivan  @lauriesullivan, June 15, 2015

It's rare to get several perspectives on any one topic in an article, especially in real time, but Mobli Media has built machine learning and computer vision algorithms that search and serve crowdsourced images and videos in articles on publisher sites. The platform, Eyeln for Publishers, announced Monday, curates content from across social sites with help from a snippet of code.

Moshe Hogeg, CEO and founder of Mobli Media, wants to reinvent search -- at least as it relates to content. 

Content continues to gain significance for search marketers because it helps pages rank higher in queries on Google, Bing, Yahoo search engines, mobile and desktop, depending on a variety of factors like authenticity, originality and timing. Irit Singer, CMO at Tel Aviv-based Mobli Media, said several publishers, such as Huffington Post, have begun to use the technology based on EyeIn, a visual search engine for user-generated content.

Mobli makes the tool available as a Web search engine through a portal, a mobile app, and as a browser plug-in. The algorithms search an event based on the metadata embedded in the image of video per locations and trends.

For instance, publishers running articles about the Apple Worldwide Developers Conference or an upcoming basketball, baseball or football game can curate images from across social networks and post them on the page to accompany the text. Mobli makes the tool available as a Web search engine through a portal, a mobile app, and as a browser plug-in. "You get the crowd's perspective in images and videos, Singer said.

Through images and videos, tools like this one from Mobli will give marketers insight into the human psyche during any moment in time. It could extend real-time targeting based on data that reflects emotions, which more platforms focusing on audiences have begun to do.

Singer -- who spent time developing algorithms before moving on to L'Oreal, Microsoft, and then Mobli--said the company plans to introduce a revenue-sharing model for EyeIn in August or September that will support some of these functions.