Home Our Success Stories Data Management for a Political SaaS Application

Data Management for a Political SaaS Application

Ajay Bidyarthy

July 27, 2023

4773

Client Background

Client: A leading tech firm in the USA

Industry Type: IT

Services: SaaS, Products

Organization Size: 100+

The Problem

As per the guidelines and discussion. Political Research Automated Data Acquisition (PRADA) in the following phases which included.

1. Get pics for existing EOs (Elected Officials)

2. Get new EOs and Pictures.

3. Run QA checks regularly on EOs

4. Get data from government Facebook pages.

5. Geospatial project: Create a new version of provided KML without using google earth. Creating a nested directories which contained description and Map-URL at the designated location.

6. Get data of US States and Counties(Including Boroughs and Parishes)

By building an automated generated structured data that allows a non – programmer to create a config for each page allowing a bot to scrap and update the data.

Our Solution

We created an automated python scripts for designated phases with respective requirements. Solutions to various type of problems varied such as most of data scrapping automation was done through python developed scripts including the geospatial KML task. In addition to this different ranges of data was scrapped generated directed output for the respective tasks in the form of CSV format. So the user’s main aim requirement was achieved i.e. a non programmer could create a con-fig and initiate a bot to scrap the required data.

Solution Architecture

The majority task of project consisted of web data scraping automation so a high- level overview, and specific implementation details of project shall will be as follows:

Web Scraping Framework: Python as a coding language was used in almost all of the tasks and the framework used for data scraping included Beautiful-Soup, Selenium and Web drivers. These libraries provide tools and functionalities to navigate web pages, extract data, and handle various HTML elements.
Data Extraction and Parsing: Use the selected web scraping library to extract the desired data from the web pages provided either in the data sheet or within the websites of URLs given in the sheet. This involves locating HTML elements, applying filters or selectors, and parsing the extracted data.
Data Processing: Followed by data extraction it was cleansed, transformed and aggregated to a structured form such as pandas’ Data Frame followed by a CSV file. In the case of geospatial task it resulted to generation of nested folders in a kml file.
Data Storage: The how and where to store the scrapped data was determined which is local file system in the form of CSV (Comma Separated Values). As it was the appropriate data storage solution according to need of the project. In addition to this the geospatial task had the output in the form of kml file as polygons inside directories of nested folders.

Deliverables

Tasks	Outputs (CSV/KML/XLSX)	Python Scripts
Canada EOs	mydata.csv	Script1.py	Script2.py
Geospatial Task	Electoral Districts.kml	–
Facebook Scrapping of EOs	EO_OUTPUT_O.csv	final_eo_scrapping.py
Facebook Scrapping of 429 Cities	Output_DRAFT_429_CITIES.csv	Facebook_image_scrapping.py
USA States Website URLs	ScreenScrapingt.csv	final_50_states_scrapping.py
USA Counties Website URLs	US Website_final_write.xlsx	county_scrapping.py

Tools used

Python (Programming Language)
Beautiful Soup
Selenium
Pandas
Numpy
Simplekml
re (regular expressions)

Language/techniques used

Python (Programming Language) – It is an interpreted language, which allows quick prototyping and interactive coding. Its versatility can be is one of the reasons for its major applications. Different libraries and tools were used in this project for various data solutions.
Beautiful Soup – A python library used for web scraping and parsing HTML and XML documents. It provides a convenient way to extract from the said files. It eases out the work flow from parsing to data extraction and encoding handling as well.
Selenium – A python library used for web browser automation like Chrome, Firefox, Safari and others. It interacts with elements such as clicking buttons, filling out forms and selecting drop down options. In this project we used it in Chrome. Selenium web driver was used for web automation. It acted as bridge between Python code and the Web browser.
Pandas – It is Python’s versatile library that provides high performance data structure tools and it is built on top of Numpy. Data Frame is one of its key feature due to which this library was used. This key feature allows efficient manipulation, slicing, and filtering of structured data
Numpy – It is also a python library aka Numerical Python as it is a fundamental library for scientific computing in Python.
Simplekml – It is a python package which enables you to generate KML with as little effort as possible.
re (Regular Expressions): It is a powerful tool in python sued for pattern matching and manipulations of strings.

Skills used

Python Programming
Web Fundamentals
Web Scrapping using libraries such as BS, Selenium.
Data Cleaning and Processing
Problem- Solving and debugging.
KML structure and handling using Python’s programming.

Databases used

None. All the structured data was in the form of either python Data Frames, CSV or Excel Sheets.

What are the technical Challenges Faced during Project Execution

Firstly some of the web URLs were not accessible because they were restricted to particular range of IPs of that region
Couldn’t fetch whole data through Beautiful Soup as it couldn’t parse whole tags.
List of US Counties wasn’t provided in the given resource links

How the Technical Challenges were Solved

Used VPN for accessing Official sites which were not generally accessible.
Used Selenium Web driver to automate the direction at URLs which fetched complete html tags of the desired webpages.
Performed a search and created structure data of list of counties of each state which was used as input to gain web URLs of counties of US.

Business Impact

Enhanced Analysis: Web scraping allows businesses to gather valuable data from various websites. This information can provide insights to desired aim and objectives enabling businesses to make informed.
Real-Time Monitoring and Upgradation: Web scraping can enable business to monitor changes or updates on website in real-time. This can be useful for tracking regulatory changes. It keeps the business and it’s data updated.
Increased Efficiency: Automation eliminated the need for manual data collection, saving time and resources. With automated web scraping, business can extract large amount data quickly, accurately, improving overall operational efficiency.