Analytics Update

Visual Analytics Searching

Related products: Explore
Visual Analytics Searching

Visual enrichments are a major leap forward in AI-powered analytics. Visual enrichments let you gain insights about your brand in non-verbal/non-textual content posted on news, blogs, and Reddit. Visual enrichments allow users the ability to search the context of an image, rather than traditionally only text for news, blogs, and Reddit documents in Explore. Visual enrichments allow users to search for specific logos, celebrities, genders, etc, via newly dedicated boolean. We offer 11 different enrichment types.

 

Learn More: Visual Analytics Boolean Operators

FAQ: Article

 

 

Really powerful stuff and a long time coming for MW! Is there a way to train the model to ID our logos for example? Our competitors’? Stuff unique to our business say when people are in front of a medicine cabinet or have their vitamins on the counter? I can’t seem to find the exhaustive list of known-logos, etc.

Love that NSFW has a image recognition capability, nsfw:true and nsfw:false. The amount of operators in Explore is second to none, really impressive how this has expanded over the last couple years. Great for power users.


Really powerful stuff and a long time coming for MW! Is there a way to train the model to ID our logos for example? Our competitors’? Stuff unique to our business say when people are in front of a medicine cabinet or have their vitamins on the counter? I can’t seem to find the exhaustive list of known-logos, etc.

Great question. Your account manager would be the best person to liaise with on this so that can gather the requirements and walk through the necessary steps with our internal teams. 


Thanks Robert, can I tag people in here? @Jess Ullrich 


@Danny Gardner 

Hey Danny, long time no speak! Glad you share our excitement around this 😀

There is a list of over 3200 logos that we support today and more are added daily. 

Yes, additional ones (your brands, competing brands etc.) can be added to the list.

If you speak to Jess, she can help with that.

The logos can be combined with objects and scenes using the relevant operators, so for example if you wanted to look for a specific logo, on a bottle, in a bathroom….that is possible!

Example:

imageEntity: "Pfizer" AND imageObject: bottle -> would give you hits of the Pfizer logo on a bottle.

Give it a try and let me know how you get on!


@john.box what’s up John! Great to see you in here, yes we’ll have to catch-up soon.

Excellent, happy to hear that, there is great potential for enhancing our passive listening and brand protection here. I’ll talk to Jess and see what we can do, thanks!!