Citizens use a variety of channels to communicate, including various Grievance Portals, Social Media Posts, Blogs, newspapers, etc. These are all unstructured & semi-structured data. The challenge lies in identifying and categorizing citizen feedback and data. This feedback is intended to understand whether citizens’ attitude and opinion towards a department, scheme, or other section are positive, negative, or neutral.
Analyzes large volumes of unstructured data by using predefined templates, machine-learning methods, and Natural Language Processing (NLP) to identify deeper insights and sentiments.
Core Sources of data Crawled & Analysed –
Data Crawling & Scrubbing
- •Data Crawling
- •Data Parsing & Filtering
- •Robust Metadata based Repository
- •Group Related Topics
- •NLP based Categorization
- •ML based new topic build
- •Sentiment tonality such as Positive, Negative, and Neutral.
- •Sentimental KPIs
- •Graphical Representation