Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/chewzzz1014/fyp-ner-archive
https://github.com/chewzzz1014/fyp-ner-archive
final-year-project flair fyp machine-learning ner nlp project spacy transformer
Last synced: about 19 hours ago
JSON representation
- Host: GitHub
- URL: https://github.com/chewzzz1014/fyp-ner-archive
- Owner: chewzzz1014
- Created: 2025-01-14T09:21:26.000Z (5 days ago)
- Default Branch: master
- Last Pushed: 2025-01-17T10:12:14.000Z (2 days ago)
- Last Synced: 2025-01-17T11:23:11.903Z (2 days ago)
- Topics: final-year-project, flair, fyp, machine-learning, ner, nlp, project, spacy, transformer
- Homepage:
- Size: 12.6 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Overview
Stored the resume dataset annotation and NER trained models for my Bachelor Degree's Final Year Project with title **AI-Powered Job Application Management for Applicant**. There are 3 main parts in the project
1. Named Entity Recognition (NER)
- Involved in complete machine-learning lifecycle: dataset annotation, model training, model evaluation
- Trained 2 NER models using spaCy and Flair library, with manually-annotated resume dataset which can recognise 11 types of entities in resume and job description.
2. Job-Resume Matching Scoring
- Calculate the matching score between a resume contents and job description in range of 0 (not relevant) to 1 inclusively (very relevant).
- Perform cosine similarity calculation between resume skills and job description skills which are extracted using NER trained models.
3. ResuMatch, Job Application Management System
- Develop job application management system consists of Job Management, Resume Management, Job-Resume Matching and Dashboard.
## Related Repositories
1. [NER model development and backend code](https://github.com/chewzzz1014/fyp)
2. [Frontend code](https://github.com/chewzzz1014/fyp-frontend)
## NER Entities
| Category | Entity | Label Name | Example |
|---------------------|-----------------------------|-------------|-----------------------------|
| **Personal Information** | Name | NAME | John Doe |
| | Location | LOC | California |
| | Phone Number | PHONE | 0192394945995 |
| | Email Address | EMAIL | [email protected] |
| **Education** | University or School Name | UNI | University of California |
| | Academic Qualification Name | DEG | Bachelor of Computer Science|
| | Study Period | STUDY PER | Oct 2021-Jul 2024 |
| **Working Experience** | Job Title | JOB | Java Developer |
| | Company or Organisation Name| COMPANY | Abc Company |
| | Working Period | WORK PER | Aug 2024-Current |
| **Skill** | Skill | SKILL | Java |
## TODO
Add completed full thesis and poster
## Spacy NER Performance
Trained Model: https://drive.google.com/drive/folders/1c8PlRGcUW96eEthZizOoK1jhYA3GjhFV?usp=sharing
```
================================== Results ==================================
TOK 100.00
NER P 83.77
NER R 85.67
NER F 84.71
SPEED 13556
=============================== NER (per type) ===============================
P R F
NAME 91.63 92.02 91.82
EMAIL 95.00 97.44 96.20
PHONE 93.67 96.73 95.17
SKILL 84.33 87.85 86.06
UNI 63.71 70.72 67.03
DEG 78.67 77.59 78.12
COMPANY 74.88 68.60 71.60
JOB 73.65 55.81 63.50
WORK PER 92.92 93.83 93.37
LOC 81.17 80.98 81.07
STUDY PER 84.96 81.70 83.30
```
## Flair NER Performance
Trained Model: https://drive.google.com/drive/folders/1CvpFy6hUTDJF6KK7XdrsYkMiKqnr2q2W?usp=sharing
```
Results:
- F-score (micro) 0.7992
- F-score (macro) 0.8069
- Accuracy 0.6678
By class:
precision recall f1-score support
SKILL 0.7980 0.8160 0.8069 20125
JOB 0.6424 0.6241 0.6331 1160
LOC 0.7402 0.8491 0.7909 782
WORK 0.8745 0.9293 0.9011 735
COMPANY 0.6924 0.6924 0.6924 699
UNI 0.7109 0.6827 0.6965 353
DEG 0.8162 0.7957 0.8058 279
STUDY 0.7729 0.8271 0.7991 214
NAME 0.8791 0.8915 0.8852 212
PHONE 0.9406 0.9596 0.9500 198
EMAIL 0.9096 0.9194 0.9144 186
micro avg 0.7900 0.8085 0.7992 24943
macro avg 0.7979 0.8170 0.8069 24943
weighted avg 0.7897 0.8085 0.7988 24943
Loss: 0.18681490421295166'
```