Machine Learning: The Driving Force of Web Scraping | OxyCast #6

A brand new episode of OxyCast is live! This time, our favorite host Augustinas Kalvis (Software Developer), and a special guest Jurijus Gorskovas (Machine Learning Engineer), delve deeper into the world of Machine Learning! 

From Jurijus' experience working with machine learning to what his "OxyBrain" team does at Oxylabs, we covered everything in the #6 episode. Watch it to understand the details of machine learning and how it can make web scraping processes more efficient. 

• Spotify

• Apple Podcasts

For starters, Jurijus explains in detail what a machine learning model is. In this episode, Augustinas gives a specific idea of what machine learning model he would like to develop, and Jurijus explains what needs to be done from scratch. They discuss a machine learning model for recognizing the price on the e-commerce website. All the essential definitions, such as "feature," are also explained during the episode.

"OxyBrain" – the team inspired by machine learning

As mentioned above, Jurijus works in one of the most mysterious teams at Oxylabs – "OxyBrain." No one knows what exact projects they work on, but when they have something to announce – everyone, who is even slightly interested in the web scraping industry, is excited about it. 

One of the innovations his team maintains is the universal e-commerce parser. It's a piece of software that takes HTML content and finds various fields, such as descriptions, price, titles, etc. During this episode, Jurijus explains that you must go through a vast amount of e-commerce pages to visualize their similarities. For example, if the title is in a specific position, you can either take a picture and detect where that particular field is or use XPath to understand the depth of that page. Once you have information about many e-commerce pages, you can see statistically that the structure is often repeating. Simply put, this data helps to improve a parser. 

From the #6 episode of OxyCast, you can expect more details regarding the parsing of e-commerce pages, powered up by machine learning.

Wrapping it up

If you want to understand better how machine learning can help you gather structured public information or how to try to develop a machine learning model yourself, the #6 episode of OxyCast has it covered. You'll also learn the critical aspects of building a machine learning model for recognizing prices on e-commerce pages. 

If you have any suggestions or questions for our podcast, don't hesitate to contact us at events@oxylabs.io, and we'll try our best to discuss them in OxyCast episodes. You can also check out other OxyCast episodes that cover different aspects of the web scraping world.