Oisin003/Financial-NLP-System

Final Year Project

What's novel

Final Year Project

Code Analysis

11 files read · 3 rounds

A full-stack financial document analysis system that extracts text from PDFs (via Tika/Tesseract), runs rule-based audit flagging (RAG classification, debt burden, margin deterioration, going concern risk) in JavaScript, and performs NER, extractive summarization with financial-domain-aware sentence

Strengths

The domain-specific audit rule engine with ablation study methodology is genuinely well-designed and tested, and the extractive summarization pipeline with explainability payloads (sentence selection reasons, compression ratios, entity/figure highlights) shows real depth in making NLP outputs interpretable for financial auditors.

Weaknesses

The financial terms list is duplicated verbatim between Python and JavaScript with only a 'keep in sync' comment, creating a maintenance hazard; the core NLP techniques (spaCy NER, TF-IDF/LDA, extractive summarization) are standard library applications rather than novel algorithms, and the rule-based audit system, while well-implemented, is essentially a keyword-matching engine rather than a learned model.

Score Breakdown

Innovation

5 (25%)

Craft

67 (35%)

Traction

6 (15%)

Scope

80 (25%)

Signal breakdown

Innovation

Not Fork+1

Code Novelty+1

Concept Novelty+1

Craft

Ci+0

Tests+8

Polish+1

Releases+0

Has License+0

Code Quality+22

Readme Quality+15

Recent Activity+7

Structure Quality+5

Commit Consistency+4

Has Dependency Mgmt+5

Traction

Forks+0

Stars+6

Hn Points+0

Watchers+0

Early Traction+0

Devto Reactions+0

Community Contribs+0

Scope

Commits+7

Languages+8

Subsystems+13

Bloat Penalty-5

Completeness+6

Contributors+5

Authored Files+15

Readme Code Match+3

Architecture Depth+7

Implementation Depth+8

Evidence

Commits

Contributors

Files

12880

Active weeks

TestsCI/CDREADMELicenseContributing

Repository

Language

Python

Stars

Forks

License

—