zxcvb2002 / sentimentanalysis

Analysis Sentiment driven Text using CoreML & CreateML

SentimentAnalysis

CoreML & CreateML Practice using Twitter Sentimental Analysis

Private Notion Link

Training Data Classification

.csv -> .txt

import os
import csv
import shutil

csv_file_path = "input path"

output_folder = "ouput path"

folders = ["Negative", "Neutral", "Irrelevant", "Positive"]

for folder in folders:
    folder_path = os.path.join(output_folder, folder)
    os.makedirs(folder_path, exist_ok=True)

with open(csv_file_path, "r") as csv_file:
    csv_reader = csv.reader(csv_file)
    next(csv_reader)
    
    for row in csv_reader:
        emotion = row[index]
        text_content = row[index]

        if emotion in folders:
            folder_path = os.path.join(output_folder, emotion)
            file_name = f"{emotion}_{len(os.listdir(folder_path)) + 1}.txt"
            file_path = os.path.join(folder_path, file_name)

            with open(file_path, "w") as txt_file:
                txt_file.write(text_content)

print("End")

Testing Data Classification

Before securing validation... (Only about 500 datasets were tested...)

Action

2023-06-22.12.16.41.mov

Reference

LinkedIn

sentimentanalysis's Issues

feature/Automation

Delete Anlysis Button and Automation Sentimental Analysis

Testing Dataset

Amount of Testing Dataset is so small.
More testing datasets are required to secure high validation.