Goodbase.AI is a reliable Chrome extension and search engine. Our experts use crawling, scraping and Natural Language Processing (NLP) to predict sustainability ratings for all fashion brands.
The Challenge: Sustainability in Fashion
In the fashion industry minimizing environmental impact, improving poor working conditions and uplifting animal welfare are unfortunately often overlooked in favour of maximizing profit. There is pressure to complete new orders every season. This holds for the modern unsustainable ‘fast-fashion’ industry which produces most of our clothes.
Many great initiatives are pushing in a different direction by selling clothing with a sustainability label. However, how do you know as consumer which brands value these beliefs? Searching for information on every single piece of clothing is very time consuming.
Would it not be great to have one platform for predicting sustainability ratings for fashion to find one reliable answer?
The Approach: Predicting Sustainability Ratings
That is where we come in. Our model is trained to predict sustainability ratings for as many clothing brands as possible, which are saved in a database for the public to see.
Our process consisted of two main steps: gathering the data (crawling and scraping) and predicting the sustainability for each of these brands. To be able to build a supervised model, we needed some examples of brands which are known to be sustainable. Firstly, we collected ratings from a few initiatives. Moreover, they manually rate clothing brands based on their level of sustainability, such as Project Cece. Finally, We use the small dataset to train an Artificial Intelligence (AI) model which provides sustainability labels for brands that are not yet rated. Here, we apply text classification, which needs a labelled dataset containing text documents and their labels to train the classifier.
The Solution: Goodbase.AI
Since we strive for real impact, the end result is not just a model and a database. We make the results available to the public both on our website, as well as via the chrome plugin. On these platforms consumers can search more than 2500 brands ranked on their sustainability. Moreover, it provides a reliable answer to the degree of sustainability of a clothing brand, before purchasing a product of that specific brand.
We encourage every brand to become more transparent and provide more information. Since our model is self-learning and dynamic, it can adjust the explainable sustainability criteria as brands provide more and better information
We created the tool for people who value our planet, people and animals, and want to live in a world that is sustainable and fair. Together we can have real impact with the one powerful weapon we all have in common: the ability to choose. That is to say, choosing and rewarding responsible fashion brands that do good, over the ones that do not, can drive the whole industry to become more sustainable and fair.
Would you like to know more about this case or what Amsterdam Data Collective can do for you? Get in touch with Casper Rutjes at firstname.lastname@example.org, or check our contactpage.