The Social Data Mining Process

Exhibit 25.2   Social Data Mining Process — authentication, data collection, data processing (cleaning and augmenting the data), data modelling and visualizing the resulting metrics.

The process of social data mining is a systematic approach that involves several critical steps (refer Exhibit 25.2):

1. Authentication: The initial step in the data mining process is authentication, where access to social media data is secured. This step is typically performed using the industry-standard Open Authorization (OAuth) process. OAuth is a three-legged process involving three key actors: the user, the consumer (which is the application seeking access), and the resource provider (the social media platform). This process ensures that only authorized applications can access user data while maintaining user privacy and security.
2. Data Collection: After successful authentication, the data collection phase begins. This step involves gathering the data that the application has been granted access to by the social media platform. The data collected can range from user profiles and posts to likes, shares, and comments. The scope of data collection is strictly confined to what the user has permitted, ensuring compliance with data privacy regulations.
3. Data Processing: Raw social media data often contains noise, such as irrelevant information, duplicates, and inconsistencies. Data cleaning and pre-processing are crucial steps to prepare the data for analysis. This involves filtering out unnecessary data, handling missing values, and transforming and augmenting the data into a suitable format for analysis.
4. Modelling (Analytics Engine): In this phase, the cleaned and processed data is analyzed using various modelling techniques. The analytics engine applies algorithms and models to identify patterns, trends, and insights from the data. This step is where the core analysis happens, utilizing techniques like sentiment analysis, classification, clustering, and more.
5. Visualization: The final step in the data mining process is the presentation of results. The insights gleaned from the analysis are presented in a format that is easy to understand and actionable. This could include visualizations, reports, or dashboards that summarize the findings and provide recommendations for decision-making.