Abstract
Companies want to change the way that users interact with their services. One of the main ways to do this is through messaging. It is well known that different users are likely to respond to different types of messages. Targeting the right message type at the right user is key to achieving successful behaviour change. This paper frames this as a case based reasoning problem. The case representation captures a summary of a user’s interactions with a company’s services over time. The case solution represents a message type that resulted in a desired change in the user’s behaviour. This paper describes this framework, how it has been tested using simulation and a short description of a test deployment.
| Original language | English |
|---|---|
| Pages (from-to) | 345-359 |
| Number of pages | 15 |
| Journal | Lecture Notes in Artificial Intelligence (Subseries of Lecture Notes in Computer Science) |
| Volume | 8765 |
| DOIs | |
| Publication status | Published - 2014 |
Keywords
- Case based reasoning
- Recommender systems
- Simulation