Machine Learning-based Fuzzy Matching of Items for Retailers and Brands
Matching items between multiple sets of data is a time consuming activity. Retailers may want to compare their products and pricing with data they scraped from competitors, but those competitive listings may be described differently. Consumer Goods companies may need to integrate data from suppliers or acquisitions.
Fuzzy matching of items is a core function in marketplaces. Online retailers may compare their listings with their competitors’ to identify differences in price and inventory. Suppliers making products across multiple sites may examine how their goods are presented to ensure consistency. Being able to link together data from different sources is a time-intensive prerequisite for operations and analytics.
In this live workshop, we will showcase our new Fuzzy Matching of Items Solution Accelerator. The accelerator is a robust analytic model with best practices and notebooks that are the foundation for a full ML-based Fuzzy Matching solution. Use machine learning and the Databricks Lakehouse Platform for product matching that can be used by marketplaces and suppliers. Resolve differences between product definitions and descriptions and determine which items are likely pairs and which are distinct across disparate data sets. And while we’ll focus on text matching in this webinar, this same accelerator can be used to match multiple types of data (e.g. images, etc.).
Together, we will cover:
- The challenge brands have with matching items at scale
- How Databricks Lakehouse helps companies rapidly perform entity resolution
- Common Use Cases and the Business Value of Fuzzy Matching for Retailers and Brands
- Deep Dive on Fuzzy Matching for Items Solution Accelerator
Rob Saker, Retail & Consumer Goods Industry Leader & RVP at Databricks will lead the workshop. Bryan Smith, Technical Director of Retail & Consumer Goods and Luke Bilbro, Solution Architect will walk through the product matching Solution Accelerator. We look forward to addressing all of your questions and hope you will join us for an interactive discussion on the topics raised during the presentations.