Skip to content

yogg17/TestPod

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

TestPod: Automated Subjective Evaluation System

Spring NLP Database

Overview

TestPod is a full-stack web application designed to bridge the gap between automated objective testing and manual subjective grading. It leverages Natural Language Processing (NLP) and Rule-based Fuzzy Logic to evaluate essay-based questions, significantly reducing the administrative burden on teaching faculty while providing students with immediate, transparent feedback.

Core Capabilities

  • Automated Evaluation Engine: Utilizes the Maximum Entropy Algorithm for POS tagging and sentence detection to analyze the semantic structure of student responses.
  • Intelligent Grading Logic: Implements a hierarchical evaluation model that maps student "answer skeletons" against faculty-provided keys using weighted average scoring.
  • Fuzzy Logic Inference: Employs rule-based fuzzy logic to handle the nuances of natural language, generating discrete scores and performance remarks.
  • Comprehensive Analytics: Offers a dedicated Faculty Dashboard featuring test statistics, student performance distributions, and step-by-step evaluation reports.

System Architecture & Workflow

  1. Ingestion: Faculty upload questions and "Gold Standard" answer keys.
  2. Decomposition: The system uses SentenceDetectorME to segment text and en-pos-maxent for Part-of-Speech tagging.
  3. Skeletonization: Answers are transformed into a <POS Tag, Token> mapping, removing noise (case conversion, trailing spaces).
  4. Comparison: The student container is cross-referenced with the faculty container to generate a Similarity Index Matrix.
  5. Scoring: The matrix is de-fuzzified and marks are calculated using a weighted average technique.

Technical Stack

Layer Technology Key Implementation
Backend Spring MVC (Java) Controller-driven architecture with Java Bean containers.
NLP Engine Apache OpenNLP Tokenization, POS Tagging, and Sentence Segmentation.
Database MySQL Relational storage for Test IDs, metadata, and user stats.
Evaluation Fuzzy Logic Rule-based remark generation and de-fuzzification.
Front-End JSP & JSTL Dynamic UI for Examination Centres and Dashboards.

Key Modules

  • Examination Centre: Secure portal for students to access tests via unique IDs and submit responses.
  • Auto-Generated Reports: High-fidelity feedback loops providing a step-by-step breakdown of how a score was derived.
  • Digital Question Paper: An optimized interface for long-form text entry and real-time submission.

Demonstration

Snapshots of TestPod

Home page

alt text for screen readers

Login

alt text for screen readers

Faculty Dashboard

alt text for screen readers

Exam Dashboard

alt text for screen readers

Auto Generated Report

alt text for screen readers

Examination Centre

alt text for screen readers

Digital Question Paper

alt text for screen readers

Developed to modernize the backbone of the education industry through scalable AI solutions.

About

TestPod is an AI-powered educational technology platform that automates the evaluation of subjective, essay-based answers. Developed using the Spring MVC framework and Apache OpenNLP, the system replaces manual grading with a high-precision NLP pipeline. By decomposing paragraphs into POS-tagged "answer skeletons" and utilizing Fuzzy Logic.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors