Pricing Assistant Presents Research at Cambridge

Exactly one month ago, on August 28th and 29th, 2015, Pricing Assistant was at Cambridge for the EuroSciPy conference. EuroSciPy is a multidisciplinary conference centered around the usage and development of Python technology in a scientific context.

The annual conference consists of two days of technical presentations and posters. Due to its flourishing ecosystem, Python is becoming more and more important in the scientific fields. Consequently, the EuroSciPy conference keeps growing and brought together more than 200 attendees this year.

A few months ago, we applied to present a poster and participate in the conference. The abstract we submitted featured our work on matching ecommerce offers. Our approach caught the attention of the conference panel and we were thrilled to be invited to present our research in two 120 minute sessions.

We created a poster presenting Pricing Assistant’s recent work on product matching. The goal was to create a tool capable of determining if two web pages are selling the same product. Our approach combines various techniques from the fields of image analysis, semantic analysis and machine learning. The technique had great results and outperformed existing literature in fields such as skincare, cycling equipment and sporting goods.

At the conference, we got the opportunity to listen to European researchers share their fascinating findings from using Python to solve major problems in fields such as astrophysics, biology and artificial intelligence. We also received some excellent feedback on our methodology, and have defined the next steps towards an even more automatic matching solution.

Here is the poster that we presented at the conference:

entity-matching-of-web-offers-from-html-to-similarity-score-1-638