1 of 19

​

​

​

​

​

Namastay,

I am Neha

​

2 of 19

  • Formerly Scrapinghub - Founded in 2010�
  • Fully remote company with 180+ employees in 28 Countries - 100% Remote since the beginning�
  • Authors and largest maintainers globally of the Scrapy open source framework�
  • Product innovators - Scrapy Cloud (2011), Smart Proxy Manager (2012), Data on Demand (2014), AutoExtract API (2019), Launched Automatic Extraction (2021), Zyte API(2022), Zyte Enterprise(2023)

A bit of

history

3 of 19

What is Web Scraping?�

​

​

​

​

​

4 of 19

What is Web Scraping?�

​

​

​

​

​

5 of 19

Next-level Personalization: How Web Scraping and Graph Databases Power Recommendation Engines.

6 of 19

  • Elements
  • Interconnection
  • Purpose/Goal/Function.

6

A System

Elements

Interconnection

Purpose/Goal/Function

Tangible and Intangible contents that work together. ��Nodes

How parts are interconnected.��Relationships

INPUT

Output

|

7 of 19

The Purpose of the Recommendation �SYSTEM �

​

​

​

​

​

DATA �IN

Recommendation System

  • Increased Engagement
  • Improved Sales/Conversions
  • Enhanced User Satisfaction��Through personalisation and enhanced discoverability.

8 of 19

  • Elements
  • Interconnection
  • Purpose/Goal/Function.

8

Traditional Recommendation System

INPUT

Output

|

9 of 19

  • Collect quality and relevant data by improving the Data Collection Process. �
  • Storing the Data in a database that offers more natural representation of the data. �
  • Initiate the design process by first articulating a purpose statement for your project, integrating relevant contextual factors specific to your use-case.

Improve

10 of 19

DATA PIPELINE�

​

​

​

​

​

11 of 19

Challenges in large Scale Scraping�

​

​

Scraping Framework

Increased Complexity & Cost

�

Proxy

Browser Automation

Anti Ban Measures

Cloud

  • Multiple Solutions
  • High Cost
  • Quality and Maintenance

12 of 19

Zyte API

Scraping Framework

Proxy

Browser Automation

Anti Ban Measures

Cloud

Scraping Framework(Scrapy)

Zyte API= Proxy + Browser Automation + Anti Ban Measures

Cloud

13 of 19

  • Elements
  • Interconnection
  • Purpose/Goal/Function.

13

Graph-Powered Recommendation System

|

14 of 19

The Purpose of the Recommendation �SYSTEM �

​

​

​

​

​

DATA �IN

Recommendation SYSTEM

GraphDb

Quality Data with tools like Zyte API

  • Increased Engagement
  • Improved Sales/Conversions
  • Enhanced User Satisfaction��Through personalisation and enhanced discoverability.

Contextual Purpose Statement

15 of 19

�Purpose Statement of a Book Recommendation System

​

​

​

​

​

"This project aims to develop a context-aware recommendation system for our online bookstore, offering dynamic, personalized book recommendations. Going beyond past behavior or explicit preferences, the system integrates contextual aspects of user interactions, such as browsing time, device, speed, and seasonal influences. With these factors, the system seeks to boost user engagement, broaden book discovery, increase sales, and provide a continuously evolving, personalized reading journey tailored to users' real-time needs."

16 of 19

Books Recommendation System �

​

​

​

​

​

Use Zyte’s expertise to scrape quality and relevant data.

​

​

​

Store it in Neo4j database.

17 of 19

�

​

​

​

​

​

Q & A

18 of 19

�

​

​

  • Sign-up for �the Zyte API.

​

​

  • Sign-up for the Newsletter

​

19 of 19

Thank you