Hullabalook for Product Data

We believe that having products that are enriched with relevant information helps shoppers discover more of your catalogue, and can help you build more exciting shopping experiences into your site.

Ultimately having enriched product data enables you to offer your customers a better filtering and search experience – getting them to the item they want to purchase faster and increasing the chance that they will convert. Through our work tagging over 1 million products for retailers we have developed a data enrichment framework for the Hullabalook platform that covers all types of products, from dresses all the way to lighbulbs

Our data enrichment & auto tagging process

1 – Unstructured Data Processing

We extract all relevant information from your product data. We use NLP techniques to understand all of your unstructured data, such as descriptions and titles.

2 – Computer Vision

We use computer vision techniques to understand and extract information from your images such as colour, pattern and length

3) Data Fusion
We fuse all of the extracted text data with all of the image data to create an enriched product

4) Hullapedia

We can standardise all of this enriched data using our proprietary Hullapedia library to create Attributes or Tags that can be fed back into your Facets/Filters through an API e.g. “Polka Dots” “Spots” “Dot” “Dotty” all become ” Dots”.

5) Modelling

We run a multi-tiered Machine Learning driven Hybrid Bayesian Classification Voting model over this enriched and standardised data to identify new types of tags such as Styles, Occasions, Trends

Why not say it with Stripes?

Hullabalook helped John Lewis to automatically select curated sets of products to be merchandised in the S/S18 seasonal trends campaigns.