Recommendation engines are the norm on websites. The problem is that the basic, statistically driven engines don’t understand an artistic point of view, especially when they relate to images, such as B. on home decor websites. While recommendation engines are great for more statistical analysis and relationships, something else is required to improve success on sites that require analytics and are more artistic concepts. Fortunately, machine learning (ML) is a wide range of solutions, and it is gradually being applied to design problems that can improve business both online and in retail.
Three years ago, I reported on how a company is using visual ML and self-tagging to help movers provide more accurate estimates to their customers. However, this only served to identify objects. The solution for moving companies did not have to worry about the apartment design, only the object and size identification for logistical purposes.
Home shopping has experienced a boom in recent years. Direct sale, cross-sell, and up-sell recommendation engines are critical to helping customers quickly find items to buy. This is not a problem for items where the relationship is simple, such as B. Piston and non-clog drain fluid. However, there are many areas where design comes into play. For example, people’s purchasing decisions for living rooms are more complex as colors, styles and more influence the purchasing decision.
The typical recommendation engine uses statistical analysis and clustering to suggest B to the most customers who like A. Art is more subtle. To add complexity, finding enough data for accurate clustering is difficult as many vendors may use multiple vendors for each product such as chair, couch, or lamp and each location.
One company looking for a solution to this challenge is Renovai. “Online shopping for home design is the perfect market for integrating design knowledge with machine learning and AI-based solutions,” said Avner Priel, co-founder and CTO of Renovai. “It is therefore impractical to derive a statistically meaningful solution for every buyer. Our system is based on industry best practices, rules and trends, as if our algorithm attended a design school”.
This is accomplished by design experts working closely with artificial intelligence (AI) experts to capture design knowledge, build a database of diagrams, and let the system use that information to make recommendations on websites.
On home design websites, creating scenes is different from many other websites. You can show a single couch, but a more impactful way to show the couch is in a living room scene with other objects. The Renovai system uses visual ML to identify the objects and analyze the relationships between them.
When a visitor sees a couch, puts it in a shopping cart, and then searches for lamps because the one in the scene isn’t the right one for the person, the system uses the knowledge in the graphical database to provide a list of options to look for sorted are the style relationship to the couch.
As more locations are added and supplier catalogs updated, the system identifies outliers from its existing knowledge. As with most ML systems, outliers are the keys that require the most human intervention. Whether it’s a newly identified object class or a question about the category to which an object belongs, the system makes an initial decision and then marks the change. The design teams can then change the graph and retrain the system. Additionally, this knowledge is used to automatically tag new items with far more information than the basics of color and size. The captured knowledge of designers contains more details to create the system model of artistic relationships between objects. The potential buyer doesn’t need to know these concepts, just that the items look good together.
Another interesting component of the solution is the support for “Click and Mortar”. There is an in-store component, delivered through a browser, where sales teams in furniture stores can use the same functionality to help customers. Home design companies, especially those with an online and physical presence, can improve their performance by addressing market needs through both sources.
I’ve always viewed programming as a craft, not as a technique or a science, but as a mixture of these and an art. Home design is also a craft. It’s interesting to see how one craft uses its craft to help another. Mixing human and machine knowledge to tackle the home design selling challenge is an interesting application of AI and other tools, a mix that can help both the businesses and the end customers.