Development of a Web Conversational Agent for a Smart Home scenario based on the task model ANSI/CEA-2018

Lorena Sancho-Mayoral. (2016). Development of a Web Conversational Agent for a Smart Home scenario based on the task model ANSI/CEA-2018. Final Career Project. ETSI Telecomunicación, Universidad Politécnica de Madrid.

Abstract:
This report is the result of a project aimed at developing a web conversation agent to guide the user in configuring devices in a smart home. This web conversation agent is developed through a design pattern called Model View Controller. This pattern allows the user to interact with the agent, through a deployed View, using JSPs or web components where the user can choose which device he wants to configure. Following this, the agent will show you the steps to follow to achieve your goal. Furthermore, we will see how these guidelines are transmitted through one to one interaction between the agent and the user. In this way the user is guided by a number of sub tasks, which are based on the development of a dialogue tree. This dialogue will automatically discard branches depending on the options chosen by the user. The answers will be presented to the user after each interaction. So, users can choose to continue with the conversation agent who chooses what branches of the tree should be followed and then presents the next options to choose. To continue, we will see how that tree dialog is presented in a hierarchical way based on the model of tasks developed in the ANSI / CEA- 2018 standard, a tool that facilitates the design of these models and which that aim to guide users in a given task. Finally, we gather the conclusions extracted from this project, the technology we have learned about during the development and possible lines for future work.