Every fledgling blogger faces the problem of what to write about or what to film - a constant pursuit to find what content would be interesting to others. It can be a struggle, and often they would fall into a loop of hits and misses. This can be a vicious cycle as they are unable to fully understand the reasons behind the success of their first publications and why the ones that follow fail to produce the same results.
For several years, a Media Research Group at ITMO’s Machine Learning Lab,has analyzed social media content using neural networks. . Now, its head professor, Aleksandr Farseev, has created a platform that would help bloggers understand their audiences better.
“Some two weeks ago, we launched the SoPop.ai beta platform that offers bloggers a detailed analysis of their publications,” comments the professor. “The free version of our app will help you analyze yourself, as well as your social media accounts, and become a blogger. You’ll come to understand who can become your audience and what kind of content you should create.”
For bloggers with experience
Advice for fledgling creators is just the tip of the iceberg. For bloggers with experience, the platform will offer guidelines for content generation, as well as assistance with finding advertisers and developing new formats.
“Having registered on the platform and stated the price rates for advertising, bloggers will be able to work with brands and communicate with companies directly,” explains Aleksandr Farseev. “They will also learn about their competitors, see the tactics they use to attract large audiences. So what we get is some kind of an advanced training course for bloggers.”
SoPop.ai is also collaborating with the Arival bank on the next stage of the platform’s development – a digital bank for influencers. In the future, bloggers registered in the system will be able to get bank services here: open accounts, receive payments, and even get blogger plastic cards of their own. According to Aleksandr Farseev, this will create an entire ecosystem for bloggers and companies that will not just help look for advertising opportunities but also improve content.
For companies
The project lets companies search for bloggers, as well. Its algorithms analyze blog content through more than eleven thousand text and visual parameters. The neural network then uses this analysis to match user profiles to different brand criterias. This makes it possible for the companies to easily find content creators that match their brands needs.
“For example, one of our clients found women in the Arab Emirates who have children under three, have posted something associated with cosmetics at least three times over the last month, and are also Muslim and wear a hijab,” says Aleksandr Fadeev. “If these women register with SoPop.ai, they will be able make money, and the company will advertise their products.”
Research basis
ITMO University scientists have a long track record for researching social media content. In 2018, they presented their results at one of the leading conferences in the field, MM’18.
Following its results, the scientists published an article that contains a description of the neural network that makes conclusions about a blogger’s personality based on their profile and publications.
