Main Idea

Why IDEA?

▪ Lot of manual effort is spent in reading through all the documents, extracting the relevant information, categorizing it into actionable information
▪ The manual process results in cost overruns, quality issues and delays.
▪ There is a real need for an automated solution to extract relevant information, identify the respective departments and highlight the summary of the documents

Highlights

▪ Read and extract data from unstructured sources through an automated process
▪ Aid in text classification, clustering and analytics using NLP, Machine learning algorithms and models
▪ Automated build of Summary and Details
▪ Extract Report Layouts
▪ Categorize the text using Machine learning models and techniques
▪ Transform into meaningful insights

High-level visualization

High-level visualization of the project.

IDEA steps

1. Source

▪ Data sources can be structured or unstructured. PDF is taken as the source here.

2. Analyze
Analyzer engine comprises of:-

▪ Reader - Read the source documents by downloading from web pages(URL) or direct uploads.
▪ Extractor - Scan the whole document or the selective pages, apply filters if any performs Regex operations.
▪ Parser - Split the text into:

Sentences, Tokens, Part of Speech Tagging
Named entity, Relationships
N-grams, Key Phrases

3. Learn

▪ Classify text using machine learning algorithms.
▪ Extract Topics
▪ Extract document Summary

4. Insights

▪ Summary views, text categories, topic charts and sentiments in CSV/XL
▪ Key phrases, Field values and Information can be viewed in the CSV format
▪ Report layouts can be downloaded as CSV/HTML format