Dandelion

A Conversational Crowd Computing Platform

What is Dandelion?

Dandelion is a conversational system (chatbot) designed to support the decision-making processes at the university.

Dandelion asks questions to students/employees of the university periodically to understand, for instance, the situation of social-distancing on the campus, students’ mental well-being, etc.

Dandelion can also help students with their daily life. For instance, students can find out what/where the next lecture is by interacting with Dandelion.


Dandelion is part of TU Delft COVID-19 Digital Campus: https://www.tudelft.nl/covid/welzijn


How does Dandelion work?






Use cases

Wellbeing survey: Dandelion can periodically send a well-being survey to users to collect their answers. The questions in the well-being survey are customizable. The time that users received the survey can also be customized and personalized.

Reporting health condition: Users can actively report their physical health condition to us via Dandelion. The questions in the physical health survey are also customizable.

Timetable service: Dandelion provides the students of TU Delft with timetable services. A natural language understanding module is integrated into Dandelion, to understand users’ intents w.r.t. asking for information about lectures or exams. Dandelion is able to acquire the timetable of upcoming lectures/exams of a specific program/year of study, and send it to users using the chat feature.

Sensing crowdedness: Dandelion will be able to gather information about crowdedness, which helps in understanding the situation in terms of whether people follow the social-distancing rule on the campus. The crowdedness sensing task can be either pushed to or triggered by the users. Users will report their locations, subjective perceptions of the degree of crowdedness, and optionally provide a photo.




Now everyone can test Dandelion

Using the following link:

https://t.me/Delft_Dandelion_Bot




Publications


Team

Alessandro Bozzon Ujwal Gadiraju Jie Yang Sihang Qiu
Carlo van der Valk Bar Lerer Orestis Kanaris Vincent Gong