One of the main reasons we created the Contextual Data Marketplace(™) was to bring partners with interesting data sets and sources to the programmatic world.
One of our partners, socialcontext.ai, epitomizes that effort and why it’s important to advertisers by offering the ability to target audiences as they read important, socially conscious news.
Socialcontext detects diversity, equity and inclusion news content using patent-pending deep learning
Socialcontext was built by journalism, advertising, and data science professors at the University of Colorado Boulder. They positively target news articles, and offer unrivaled precision and safety, at no additional cost.
As such, Socialcontext allows advertisers to connect with audiences that care about important social issues. By advertising on DEI news, they can use advertising spend to reach corporate social responsibility goals. Each algorithm that Socialcontext builds is designed to identify quality, credible journalism that brings awareness to positive social change.
Now is the time to take advantage of these targeting abilities. Here’s why.
With Socialcontext, every word of every article is considered on balance before a label is generated.
Here’s how it works.
Socialcontext expands the programmatic reach of advertisers and recovers 20% of impressions blocked due to overbroad brand safety. A phrase or two in a news article with hundreds of words should not be the reason an article is deemed “safe” or “unsafe.” Socialcontext’s neural network algorithms understand the connections of the English language generally, and learn entire lexicons for each D&I category detected.
|
Content type |
% of all news |
| Climate action | 5% | |
| Pro-racial justice | 3% | |
| Female sports | 1% | |
| Gender equality | 5% | |
| Latinx & Hispanic culture | 15% | |
| LGBTQIA+ | 3% | |
| Mental health awareness | 6% | |
| Renewable energy | 1% | |
| All categories | 30% |
For instance, the LQBTQIA+ segment identifies news that celebrates persons who identify as lesbian, gay, bisexual, trans, queer, intersex or asexual and also includes news that discusses protections and rights for LGBTQIA+ persons. Additionally, it targets articles such as positive profiles of LGBTQIA+ persons and news related to inclusive corporations. Socialcontext promises never to include negative political coverage or hate-related stories.
“We created these tools so leading brands can invest in the good news that their customers and employees care about."
Chris J. Vargo, Ph.D., founder of Socialcontext and professor of advertising at the University of Colorado Boulder.
The result is a big picture understanding of each D&I concept that no other contextual service provides. At least 30% of news from the major publishers studied here fits one of our contextual categories.
And they’re not stopping there. Human experts at Socialcontext create new D&I contextual categories all the time.
Each Socialcontext segment is the product of tens of thousands of human annotations on a wide variety of news.
Algorithms are only released when its classification performance metrics reach academic peer-review standards. The result is article-level news intelligence with accuracy, precision and recall, unmatched by current sentiment and entity recognition approaches.
Ethics and transparency are at the heart of Socialcontext and its AI is ethically reviewed by outside academics. All Socialcontext users are invited to see “under the hood” to understand exactly what news is inside each curated segment.