Natural Language Processing
As a city council, the client was receiving complaints through a number of mediums including form fills, queries and emails. The council wanted a way to track the sentiment of the complaints so they could be broken down and distributed to the appropriate council departments and services.
Blackbook trained a NLP model based on the supplied training data set that was able to classify the ‘documents’ into the council’s nominated service areas. The documents were then analysed for sentiment with any key entities and phrases extracted for deeper insights. This data was presented in a visualisation layer through Microsoft Power BI.
The solution provided a quick, high level view of how the council was performing over time, with the dashboard giving a insight that users could see trends over time and drill down into areas of interest. The council could address any sudden deviations in sentiment if required.