Product Data Science: Matching Retail Products at Scale - Indix



Product Data Science: Matching Retail Products at Scale

HasGeek is a technology company in the media business based in India. They are committed to creating discussion spaces for geeks and encouraging dialog and exchange of ideas through meetups, conferences, hackathons, and other such events. At Indix, we love engaging with the community. The Fifth Elephant (organized by HasGeek) is India’s premier machine learning and analytics conference.


Nikhil Ketkar at the Fifth Elephant

This year at the Fifth Elephant, Indix Director of Engineering Nikhil Ketkar gave a talk on the challenging yet fascinating problem of matching retail products at scale. Here is a summary of the problem and a video of the talk itself. Enjoy!

Indix is a Product Intelligence company building the world’s broadest and deepest product database. Matching identical products from different retail websites is one of the hardest and the most impactful problems in the space of Product Intelligence. This talk will cover the breadth of algorithms and models we use for matching products across customer catalogs. It will also cover some practical aspects of taking these algorithms and models to production.

Product matching is the problem of resolving product entities across e-commerce sites. This involves a complex sequence of tasks which include:

  1.  automatic extraction of key information regions from raw HTML (for example, product titles, UPCs etc.)
  2. categorizing products into a unified taxonomy
  3. semantic parsing of product titles and specifications
  4. standardization of attributes such as brands, colors etc.,
  5. grouping products into clusters of matched products based on a similarity function or inferencing model. This is a challenging problem because unique and universally agreed upon identifiers are not always available and product details are noisy and often sparse. So we have to develop contextual understanding of product specifications, which are often expressed differently by retailers, merchants, aggregators etc.

To scale the matching problem to half a billion products, we also need to prune and bucket effectively while achieving good recall. Matches need to be highly precise since customers may use them for sensitive tasks such as price comparison, competitive analysis and catalog enrichment. We employ an ensemble of online and offline algorithms and models to perform matching at scale for a large number of stores, categories and brands.

Nikhil Ketkar leads the data science team at Indix ( which does the research and development around product categorization, standardization, matching, search relevance and ranking. He brings along a decade of experience in making data-driven decisions and building machine learning models.

This video has been reproduced with permission from HasGeek TV.

Below is a gallery of the slides should you wish to review them.

Also published on Medium.

Get API Access Today

Leave a Reply

Your email address will not be published. Required fields are marked *