Millions of people use and interact with Pinterest. This interaction generates data that Pinterest can use in turn to make search better. The process is a positive feedback loop. Better data makes for better personalization, more relevance.
To better understand the innovative ideas and projects at Pinterest, I reviewed the information at Pinterest Labs. I am a 'visual' person and like to see information mapped out as a figure or a graph, as is with much of my company assessments, I generated a unsupervised network map of projects Pinterest is involved in.
The 'engine under the hood' is involved in user modeling and recommendations this is done mostly through Machine learning algorithms to understand the users behavior as well as product/image relevance.
 |
| Pinterest: A visual discovery, engagement, shopping and advertisement engine (click to zoom) |
Pinterest is researching visual discovery. Pinterest Lens allows users
to get recommendations from real world objects. The company has some of
the largest datasets of visual objects. The focus is now to improve
recognition and visual search. It is also looking at understanding video
and moving images. Additional innovations in real world visual discovery, take search, engagement, advertisement and shopping to the next level.
With the troves of data that Pinterest generates, they are able to take a
systematic approach to explore trends and track changes. This allows
pinterest to be ahead of the curve so to speak in knowing what trends
and tastes are evolving. This is crucial information for advertisers. Pinterest is in effect a recommendation service. The massive troves of data leads it be one of largest-scale recommender systems. A very important statement caught my eye,
"Right recommendation to the right person at the right moment"Beside the operational side of the platform, the business side will also see significant growth. The platform is visual discovery, shopping engagement, social, search, marketing platform all in one package. Pinterest with its advanced data science, visual search analysis and a 300 billion and growing repository of human curated ideas will be able to do it!
Disclaimer: I am an investor in Pinterest (PINS). The data were collected from publicly available sources online and from company websites.
reference: https://www.pinterestlabs.com/projects/
