This is a resume parser application built using Streamlit, Spacy, Plotly, and Google Serp API. It helps in identifying the job title, skills, and experience of candidates from their resumes. Additionally, it provides a job search feature based on the job title and location using Google Serp API.

Follow the steps below to set up and run the application:

Clone this repository to your local machine using the following command:

git clone <repository-url>

switch to the main directory

cd Resume-Reader
pip install -r requirements.txt

streamlit run main.py

We have fine-tuned two models namely roberta-base for Named Entity Extraction and bert-base for generation of job title after getting the skills as keywords in input.

This files defines a set of functions and configurations for parsing and processing text data from various formats such as PDF, DOCX, and HTML. Below is an explanation of each component:

1. parse_pdf(pdf_content):

   • Uses PyMuPDF (fitz) library to parse text content from a PDF file.
   • It iterates through each page of the PDF document, extracts text, and concatenates it.

2. extract_text_from_resume(html_content):

   • Parses HTML content using BeautifulSoup.
   • Extracts text from HTML elements and preprocesses it by removing extra whitespaces.

3. parse_docx(docx_content):

   • Utilizes the python-docx library to extract text content from a DOCX file.
   • It reads each paragraph from the document and concatenates them into a single text string.

1. preprocess_text(text):

   • Removes extra whitespaces from the input text.

2. process_text(text):

   • Splits the text into sentences based on full stops.
   • Concatenates the sentences with a newline after each.

• colors: Defines a dictionary containing colors for different named entities.
• options: Configuration for entity visualization using spaCy's displacy module.
  • Specifies named entities (e.g., "JOB TITLE", "SKILLS") and their associated colors.

• PyMuPDF (fitz): For PDF parsing.
• BeautifulSoup (bs4): For HTML parsing.
• python-docx: For parsing DOCX files.
• spacy: For named entity recognition and visualization.
• plotly: For generating interactive visualizations like graphs and charts.

This code defines a Streamlit application for parsing and visualizing resume data, as well as performing job searches based on extracted information. Below is an explanation of each component:

• streamlit: Framework for building interactive web applications.
• spacy: For named entity recognition and visualization.
• plotly: For generating interactive visualizations like graphs and charts.
• pandas: For data manipulation and analysis.

1. Introduction Page:

   • Displays a welcome message and an about section in the sidebar.
   • Provides a file uploader to upload resumes in PDF, DOCX, or HTML format.
   • Processes the uploaded file to extract entities (Job Title, Skills, etc.).
   • Visualizes the extracted entities using spaCy's displacy module.

2. Process Uploaded File:

   • Parses the uploaded file based on its format (PDF, DOCX, HTML).
   • Utilizes spaCy for named entity recognition on the parsed text.
   • Renders the parsed text and named entities using Streamlit components.

3. Visualization Page:

   • Displays visualizations of extracted entities from the resume.
   • Generates a sunburst chart using Plotly Express to visualize entity categories and values.

4. Search Jobs Page:

   • Allows users to search for job opportunities based on extracted skills or job titles.
   • Users can enter a query (job title or keyword) and location for the job search.
   • Performs a job search using predefined functions from the 'jobs' module.
   • Logs the job search event.

• File Logging: Logs events such as resume upload, entity extraction, visualization, and job searches to a log file named 'resume_parser.log'.
• Provides information about the time taken for model loading and inference.