Improving the Relevance Score of News Feed Items to Personalize Social Media Content
Consistent user feedback led a global company to decide it needed to improve the system it used to create a relevance score for items in user news feeds and therefore improve personalization and engagement. To facilitate that, the company’s algorithm required an accurate representation of their user base. The company was looking for a vendor to help them expand into new markets.
The company came to Appen with a subjective and complex task. A pilot with 500 raters began. Appen developed a strong, scalable onboarding module and scoring system so raters could score the importance and impact of content.
With a strong partnership formed between the client and Appen, the 4-week pilot became an ongoing program. Not only is their news feed content much more personalized, but the client can now apply similar processes to address relevance score issues in other areas of their site, including SPAM.