omotuno / hr-analytics

Explore HR Analytics using SQL and Python for dataset creation, cleaning, and EDA. Visualize insights with Python (matplotlib, seaborn) and Tableau. Enhance HR decision-making with comprehensive analysis.

HR-Analytics

Every organization relies on its HR department for recruitment and placement-related tasks. This project focuses on leveraging a substantial dataset to extract insights that can aid HR departments in understanding the market's recruitment landscape. The report delves into various insights obtained through detailed Exploratory Data Analysis (EDA) and data cleaning processes. Explore HR Analytics using SQL and Python for dataset creation, cleaning, and EDA. Visualize insights with Python (matplotlib, seaborn) and Tableau. Enhance HR decision-making with comprehensive analysis.

Dataset Overview

The cleaned dataset for this project can be found here.

The dataset comprises job details within the analytics industry, focusing on Hong Kong-based companies. It includes diverse job roles such as Data Scientist, Data Analyst, Business Intelligence, Machine Learning Engineer, and Data Engineer. The dataset is extensive, encompassing six separate datasets merged for comprehensive analysis.

Field	Description
Job Title	Names of various job roles offered by companies.
Company	Names of companies providing different job roles.
Salary	Salary information for various job roles.
Job Description	Detailed descriptions of job roles.
Career Level	Categorization based on experience levels.
Qualification	Educational qualifications required for jobs.
Years of Experience	Minimum experience required for various roles.
Job Type	Classification of jobs (full time, internship, part time, contract).
Company Website	Links for job applications.
Industry	Domains to which job roles belong (Finance, HR, IT, etc.).
Search Term	Analytics domain classification (Data Science, Data Analytics, BI, ML, etc.).
Date Posted	Date when a job was posted.
Date Scrapped	Date when the dataset was obtained.

• The python Jupyter Notebook for this project can be found here
• The SQL file is available mysql.

Steps for EDA and Data Cleaning

• Removing Unnecessary Columns [lot of nulls and not useful for visualization and for getting the answer to business questions and probelems]

• Eliminate columns that do not contribute to the analysis, enhancing the dataset's relevance.

• Removing Lower and Upper Salary Ranges: Address irrelevant salary range columns to streamline the dataset.

• Removing Date Source Column:

• Eliminate redundant columns like the date source to maintain data integrity.

Business Questions

Business Questions and Objectives	Description
Total Jobs and Companies:
Total number of jobs available	Determine the overall count of available job positions.
Total Number of Companies Providing Jobs	Find the total count of distinct companies offering job positions.
Domain-wise Insights and Distribution:
Total Number of Industries (Domains)	Identify the total count of distinct industries or domains.
Domain with the highest number of jobs	Explore the domain with the highest job openings.
Distribution of job types across various domains	Analyze how job types are distributed across different domains.
Career Levels and Distribution Analysis:
Distribution of job types across different career levels	Understand how job types are distributed across career levels.
Job Role Distribution Analysis:
Distribution of job roles/titles across various analytics fields	Examine the distribution of job roles across analytics fields.
Minimum Required Qualifications for Job Roles:
Minimum required qualifications for job roles	Determine the minimum qualifications required for each job role.
Job Type Distribution Analysis:
Distribution of job types across analytics fields	Explore the distribution of job types across different analytics fields.
Job Types for Different Job Titles	Examine the variety of job types associated with different job titles.
Company-wise Insights & Analysis:
Companies with the highest number of job openings	Identify companies that have the highest number of job openings.
Top 5 Companies and Total Jobs	Rank and showcase the top five companies with the most jobs.

Insights & Findings

Domain Wise Distribution Analysis & Insights:

• The Total Number of Industries = 51
  

• The Industry with the highest number of jobs = Human Resources Management/Consultancy (777)
  

Career Level Distribution Analysis & Insights:

Entry Career level have the most job openings and opportunities followed by Middle Career Level then Senior Career Levels and intuitively this makes sense

Career Level	Total Job Count
Entry	1328
Middle	1000
Senior	883

The minimum required qualification for respective analytics field is shown below:

Search Term	Minimum Qualifications
Business Intelligence	(N/A)
Data Analyst	(N/A)
Data Engineer	(N/A)
Data Scientist	(N/A)
Machine Learning Engineer	Degree

The Percentage of Jobs across various career levels:

Career Level	Percentage
Entry	41.36
Middle	31.14
Senior	27.50

Analytics Field, Job Role/Title & Job Type Distribution Analysis & Insights:

• The total number of Unique Jobs Available = 2789
  

• The total number of jobs available is: 3211
  

• The total number of Analytics field = 5 {Data Analyst,Business Intelligence, Data Scientist, Data Engineer, Machine Learning Engineer}
  

The total number of jobs available across the 5 analytics field :

Analytics Field	Counts
Data Scientist	2490
Data Analyst	502
Business Intelligence	137
Data Engineer	55
Machine Learning Engineer	27

Company Wise Insights and Analysis

• The total number of Companies providing jobs = 1232
  

The top Five Companies providing jobs are

Company	Job Openings
IT Solutions Limited.	43
Robert Walters (HK)	40
Hong Kong Applied Science and Technology Research Institute Company Limited	37
Manpower Services (Hong Kong) Limited	31
Pinpoint Asia Limited	29

End to End Project Workflow

This project involved a comprehensive approach to data handling, combining the power of MySQL, DataGrip, Python, and Tableau. The process can be summarized in the following steps:

• Data Import and Database Setup The initial step included importing raw data into a MySQL database. A table was created in MySQL to structure and organize the data. This facilitated efficient querying and analysis.

• MySQL and DataGrip DataGrip was employed as the SQL IDE to interact with the MySQL database. SQL queries were executed to explore and understand the dataset. The capability of DataGrip to visualize query results in a tabular format greatly facilitated data exploration and initial insights.

• Python for Data Cleaning Python, along with the Pandas library, played a pivotal role in data cleaning and preparation. The dataset was loaded into a Pandas DataFrame, allowing for extensive data manipulation, handling missing values, transforming data types, and other preprocessing steps.

• Data Visualization

• Tableau was chosen for its robust data visualization capabilities. Cleaned and processed data from Python was connected to Tableau for creating insightful and interactive visualizations. Dashboards were designed to present key findings and trends.

• Python (Matplotlib, Seaborn) In addition to Tableau, Python's data visualization libraries, including Matplotlib and Seaborn, were employed for exploratory data analysis. Python scripts generated various charts and graphs to unveil patterns within the data.

• Project Optimization and Future Metrics

While the current project has provided valuable insights and was majorly tailored towards answering the business problems, there is always room for optimization and the extraction of further metrics. Future iterations could explore additional aspects of the dataset, leverage advanced statistical analyses, and enhance visualization techniques for even deeper insights. The visualization could really be improved on as well and maybe create a story in tableau.

Feel free to adapt and expand upon this section based on the specifics of your project.