Hello folks! I have completed a simple end to end fake job post classification machine learning system.
The dataset is a open source dataset from kaggle. The dataset consist of features like company-profile,benefits,salary,company logo, employment type etc.
The project required various libraries including basic natural language processing using NLTK, Gensim and LSA techniques for text preprocessing. Seaborn and matplotlib for visualization. Imblearn and Scikit-learn for data preprocessing and model building.
Metrics considered for this problem is recall score, as we want to have accurate fake class prediction.
The text preprocessing can also be completed using transformers libraries using HuggingFace API. Due to device limitations we could not execute the code, however the codes are provided in the file for reference.
The project is divided in 3 files.
> Visualization.
> Preprocessing.
> Model Building.
As the final stage the project is concluded by deploying simple prototype with streamlit. We have integrated Docker and streamlit for deploying process.
All this files are available in the project repository.
Thanks and feedbacks appreciated.
About
This is is simple end to end project including Natural Language Processing,Anomaly detection and Machine Learning