How to Analyze Social Media Sentiment Using Brandwatch API in Python

Set Up the Environment

• Ensure you have Python installed on your machine. Most systems come with Python pre-installed.
• Install necessary Python packages. Libraries like `requests` for API calls and `pandas` for data manipulation are crucial.
• Use the command line to install these packages with `pip`:

pip install requests pandas

Authenticate with Brandwatch API

• Brandwatch requires authentication via OAuth. You'll need your client ID and client secret.
• Perform OAuth authentication to retrieve an access token. Use the POST request to exchange credentials for a token:

import requests

def get_access_token(client_id, client_secret, url='https://api.brandwatch.com/oauth/token'):
    data = {
        'grant_type': 'client_credentials',
        'client_id': client_id,
        'client_secret': client_secret,
    }
    response = requests.post(url, data=data)
    response.raise_for_status()
    return response.json()['access_token']

Retrieve Data from Brandwatch

• Use the access token to make requests to the Brandwatch APIs.
• To fetch social media mentions, define a function that sends a GET request with the token:

def get_mentions(project_id, query_id, access_token, url='https://api.brandwatch.com/projects'):
    headers = {
        'Authorization': f'Bearer {access_token}'
    }
    response = requests.get(f'{url}/{project_id}/queries/{query_id}/data/mentions', headers=headers)
    response.raise_for_status()
    return response.json()

Process the Retrieved Data

• Convert the JSON data into a `pandas` DataFrame for easy manipulation and analysis.
• Inspect the DataFrame to understand the structure and available sentiment scores.

import pandas as pd

def process_mentions(mentions_data):
    mentions = mentions_data['results']
    df = pd.DataFrame(mentions)
    return df

Analyze Sentiment

• Identify sentiment-related fields, such as positive, neutral, and negative mentions.
• Calculate sentiment scores or ratios to evaluate overall sentiment trends:

def analyze_sentiment(df):
    sentiment_counts = df['sentiment'].value_counts(normalize=True) * 100
    print("Sentiment Distribution (%):")
    print(sentiment_counts)
    return sentiment_counts

Visualize Sentiment Data

• Utilize visualization libraries such as `matplotlib` or `seaborn` to create visual representations of sentiment analysis.
• Develop charts to highlight positive, neutral, and negative sentiment proportions:

import matplotlib.pyplot as plt

def plot_sentiment(sentiment_counts):
    sentiment_counts.plot(kind='bar', color=['green', 'grey', 'red'])
    plt.title('Sentiment Analysis')
    plt.xlabel('Sentiment')
    plt.ylabel('Percentage')
    plt.show()

Automate and Scale the Solution

• Encapsulate the process in functions or scripts to run regularly for updated sentiment insights.
• Consider using automation tools or scheduling scripts like `cron` jobs for periodic execution.

# Define main function to control flow execution

def main(client_id, client_secret, project_id, query_id):
    token = get_access_token(client_id, client_secret)
    mentions_data = get_mentions(project_id, query_id, token)
    df = process_mentions(mentions_data)
    sentiment_counts = analyze_sentiment(df)
    plot_sentiment(sentiment_counts)

# Execute the script
# main('your_client_id', 'your_client_secret', 'your_project_id', 'your_query_id')