Raw Agent Notes and Call Metrics Dataset
Data Science and Analytics
Tags and Keywords
Trusted By




"No reviews yet"
Free
About
Telecom customer service interaction data focusing on agent notes and assigned topics. Derived from experiences with popular telecom companies, this corpus addresses the scarcity of public conversational data in the sector due to privacy and reputation concerns. It primarily features raw text notes scribbled by agents during calls, capturing casual shorthands and potential spelling errors, alongside assigned issue topics, call duration, and location identifiers. This dataset is valuable for developing conversational AI and improving interactive voice response systems.
Columns
- RecordID: Primary identifier for the specific interaction record.
- CustomerInteractionRawText: The notes manually written by the agent while speaking with the customer, often containing casual shorthands and spelling mistakes.
- AgentAssignedTopic: The category of the issue assigned by the agent from a list of topics (e.g., Checking IMEI status, Port Out); subject to potential assignment error.
- LocationID: A numeric representation indicating the customer's location.
- CallDurationSeconds: The duration of the call measured in seconds.
- AgentID: Unique identifier for the customer service agent handling the call.
- CustomerID: Unique identifier for the customer involved in the interaction.
Distribution
- Format: CSV
- Size: 10.25 kB
- Structure: 7 columns, 103 records (100% valid entries for key columns like RecordID and CustomerInteractionRawText).
Usage
- Training Conversational AI and chatbots for customer support contexts.
- Enhancing Interactive Voice Response (IVR) containment flow.
- Multiclass classification of customer issues based on raw text input.
- Analysing agent note-taking patterns and shorthand usage in a telecom setting.
Coverage
- Geographic: Contains anonymised location data (LocationID ranges).
- Demographic: Telecom service subscribers and customer support agents.
- Scope: Covers interactions related to service enquiries, technical status checks (e.g., IMEI), and porting requests.
License
CC0: Public Domain
Who Can Use It
- Data Scientists developing Natural Language Processing models for customer service.
- Telecom Business Analysts monitoring common support topics and agent performance metrics.
- Machine Learning Engineers working on text classification and sentiment analysis.
Dataset Name Suggestions
- Telecom Agent-Customer Interaction Notes
- Customer Service Conversational Topics Corpus
- Telecom Support Call Logs and Topics
- Raw Agent Notes and Call Metrics Dataset
Attributes
Original Data Source: Raw Agent Notes and Call Metrics Dataset
Loading...
