An ontology and a taste graph to understand people’s taste

We are building the infrastructure to automate outfit advice, creating a personal fashion stylist for each individual. Our first objective has been to understand the taste, behaviour and needs of each individual. On top of this understanding, everything else is built.

Our enabling infrastructure contains five assets:

1. A consumer app:

Our consumer apps bring unique impact to our learning process, allowing us to see the future with more clarity, by solving the problem end to end, and receiving feedback from many people. We are strong believers in fashion-oriented hardware, and foresee a future where in-bedroom smart mirrors will be big drivers of voice commerce.

An Outfit Ideas Alexa Skill with a smart closet, an Action for Google Assistant, an Android and an iOS app regularly featured by Apple in 140 countries. Users store their clothes in their Chicisimo smart closet, receive outfit ideas for those clothes, and express their behaviour in several ways. Read more below;

2.- An ontology:

Our ontology is the classification of descriptors needed to define an outfit, in the terms that are relevant for women, using the vocabulary they use, with an understanding of how these descriptors are used together. This allows us to understand input, and respond with the correct output. We want to help people decide “how to wear her black dress to go to her friend’s wedding on a cold day”. Or help match “her specific skirt with her other clothes, considering her style”.

Chicisimo's infrastructure can process data from any source, and match third party taxonomies with our ontology. By processing 3rd party data, we can automatically create taste profiles of the corresponding shoppers. This ontology transforms any incoming data into clean and structured data, so any algorithm can make use of it (before the machine can process input, it must first understand it);

3. A taste graph:

Our taste graph allows us to understand people’s behaviour and taste. The objective of the taste graph is to deliver relevant suggestions to people. Relevancy depends on the context.

Our taste graph is composed of three elements: ontology descriptors, described outfits and people’s taste profiles. It receives data and correlations and assigns the correct descriptors to outfits and to people. Data might come from people through our Chicisimo apps, or might be provided by any ecommerce. Read more below;

4. A dataset:

The taste graph and the ontology produce a dataset of clean, structured and correlated descriptors, outfits and people. The taste graph and the ontology are key assets for machine learning. The dataset is exposed to the team via an internal dataportal, which provides us with transparency and control;

5. Patents:

We’ve patented the above system, and also a system to tag fashion images with shoppable products. An independent review of Chicisimo’s portfolio uncovered market adoption related to linking user-submitted fashion images to shoppable items. Chicisimo’s patents are expected to provide a competitive advantage, and we continue pursuing patents related to closet personalization; fashion trend detection; group clothing recommendations; and fashion-driven social networking. The Chicisimo team sold taste capturing patents in the music space to Apple in 2012. Read more here.

Our posts that you might like:

- In-store outfit recommender: Matching the clothes in your closet with garments you are about to buy;

- Taste graphs: How to understand fashion taste like Spotify does with music;

- How we grew from 0 to 4 million women on our fashion app, with a vertical machine learning approach;

- Read Apple’s description of Chicisimo as “your dedicated personal stylist.

Thanks for reading!

The Social Fashion Graph

Our taste graph is an assistant that knows your needs, and is able to respond to them.

The Social Fashion Graph is the name we’ve given to our taste graph. It processes people’s data received from our app, or from any 3rd party such as an in-commerce. The backbone of the Social Fashion Graph is our ontology, which plays a critical role at interpreting input and giving structure to the incoming data.

Taste graphs will transform fashion, we believe. They will focus on understanding post-purchase clothing behaviour, and will allow tech companies to understand taste of each individual, as Spotify does with music. Understanding taste of each shopper will allow ecommerce companies to build many ecommerce functionalities on top of that taste. And they will end up owning people’s attention, because they will be useful.

We built our taste graph to respond to some relevant questions, such as: How do people describe their clothes, outfits and what-to-wear needs? What clothes do they have in their closets and how do they wear them? What style do they like?

When we get dressed in the mornings, we establish correlations among clothes, and among our ways of describing our outfits and needs. Taste graphs, with the correct ontology, capture those correlations among descriptors, outfits and people.

Learn more about our taste graph here.

Our consumer app

Chicisimo’s iPhone and Android apps, our Outfit Ideas Alexa Skill for the Echo Show (video), and our Action for Google Assistant, connect the Social Fashion Graph to the reality of people’s needs, and captures clothing data..

Most important of all, the apps help us learn, they bring unique impact to our learning process. Thanks to our apps and skill, we receive daily direct feedback from many people, which helps us learn.

We think this is the most interesting aspect of building a consumer product. The fact that, regularly, we access new corpuses of knowledge that we did not have before. This new knowledge helps us improve the tech and product significantly, and it is a great reminder that we are not in the upper part of the learning curve -we are simply moving up.

When we’ve obtained these game-changing learnings, it’s always been by focusing on two aspects: how people relate to the problem, and how people relate to solutions.

Iterating a consumer app is a unique learning experience, contributing greatly to the taste graph and of course the ontology.

And finally, this is just a test for an outfit maker app, an outfit app, a closet app, an outfit planner app, or a wardrobe app.

We use our own and third party cookies to offer you a better user experience. When you use our services, we understand that you accept the use we make of cookies.