R Shiny dashboard for sentiment analysis and SEBI-style fraud risk screening
An interactive R Shiny dashboard for Sentiment Analysis and SEBI / SEC-style Fraud Risk Screening using Natural Language Processing (NLP).
Developed by Parmesh Kumar, this project demonstrates how textual narratives such as director speeches and company reports can be analyzed to identify sentiment patterns, narrative inconsistencies, and potential governance risk signals.
👉 Access the live Shiny app here:
🔗 https://parmeshkumar.shinyapps.io/sentiment-sebi-fraud-dashboard/
- Upload
.txtfiles - Displays the complete input text
- Calculates:
- Average sentiment score
- Sentiment percentage (0–100%)
- Overall sentiment category:
- Extremely Negative
- Negative
- Neutral
- Positive
- Extremely Positive
- Analyzes narrative divergence between:
- Director / CEO speech
- Company final or annual report
- Detects:
- Sentiment divergence
- Linguistic deception markers
- Optimism bias
- Generates a Fraud Risk Score (0–100) with categories:
- LOW RISK
- RISK
- MODERATE RISK
- HIGH RISK
- SUPER RISK
⚠️ This is a risk-screening model, not a fraud-detection or legal judgment system.
- Upload the text file
-
To perform SEBI Fraud Risk Analysis: -the text file must contain : -[SPEECH] -Director or CEO speech text.
-[REPORT] -Company annual or final report text.
- Select the option to feed text manually.
https://parmeshkumar.shinyapps.io/sentiment-sebi-fraud-dashboard
- install.packages(c("shiny", "tm", "sentimentr", "stringr", "DT"))
- setwd("path/to/Sentiment-SEBI-Fraud-Risk-Shiny-Dashboard"
- shiny::runApp()
This project is created and maintained by Parmesh Kumar.
📄 Read more about the author here:
👉 Author.md

