Professional Documents
Culture Documents
Drugs Recommended System2
Drugs Recommended System2
This Project intends to present a drug recommender system that can drastically
reduce specialist’s heap. In this research, we build a medicine recommendation
system that uses patient reviews to predict the sentiment using various
vectorization processes like Bow, TF-IDF, Word2Vec, and Manual Feature
Analysis, which can help recommend the top drug for a given disease by
different classification algorithms.
The predicted sentiments were evaluated by precision, recall, f1score,
accuracy, and AUC score.
ADVANTAGES:
The results show that classifier Multilayer Perception using BOW(BAG OF
WORDS) vectorization outperforms all other models with 99% accuracy.
SYSTEM SPECIFICATIONS
HARDWARE REQUIREMENTS
SOFTWARE REQUIREMENTS
HARDWARE REQUIREMENTS
UML Diagrams
System Architecture
Modules
UML
(UNIFIED MODEL LANGUAGE)
DIAGRAMS
SEQUENCE DIAGRAM
Read & Pre-process Dataset: using this module we will read all reviews, drug
name and ratings from dataset and form a features array.
Comparison Graph: using this module we will plot accuracy graph of each algorithm
Recommend Drug from Test Data: using this module we will upload disease name
test data and then ML will predict drug name and ratings.
Upload Drug Review Dataset
To run project double click on ‘run.bat’ file to get below screen
In above screen click on ‘Upload Drug Review Dataset’ button to upload
dataset to application and to get below screen
Read & Pre-process Dataset
TF-IDF Features Extraction
Train Machine Learning Algorithms
.
Comparison Graph
Recommend Drug from Test Data
The overall flow of the drug
recommendation system
CONCLUSION
Reviews are becoming an essential part of our daily lives; before going shopping, making an online
purchase, or visiting a restaurant, we first read reviews to help us make the best choices.
A recommender system was developed using a variety of machine learning classifiers, including
Logistic Regression, Multilayer Perception, Multinomial Naive Bayes , Ridge classifier, Stochastic
Gradient Descent, Linear SVC, applied on BOW, TF-IDF .
Using five distinct metrics—precision, recall, f1score, accuracy, and AUC score—we assessed them
and found that the Multilayer Perception on BOW performs 99% better than the other models.