Sprint 3 – Prototype - AgileBusinessAnalysis/01_TEAM GitHub Wiki

Recommendation concept

According to the sprint goal, the focus of this sprint's prototyping lays on content (questions and tipps) with the recommendation concept – answer of questions X and answer of question Y results in tipp Z – as its center.

Recommendation logic

The recommendation logic was defined during a 30 minutes workshop within the team. In the workshop, the team analysed the questions and defined possible conclusion an answer makes about the participant's partner. The result was a decision table that shows which answer combination results in which recommendation (tipp).

The objective was not to design a perfect recommendation, it was rather to set the basis for the following validation. The validation should "train" the system (without AI) so that the recommendation concept can be improved based on the participant's choices.

Recommendation concept prototype

For testing the recommendation concept a new prototype engine was needed. The used engine (Ionic Creator) is not able to deal with the needed logic. The objective of the prototype is to verify the quality of the recommendation logic, thus no effort was put into the design of the prototype. This led the team to choose a simple questionnaire framework: interact. The resulting recommendation concept prototype looks as follows.

First, the participants are greeted by a welcome screen which explains the basics.

Then, a series of three questions about the participant's partner are asked – similar to the prototype of the first two sprints. The questions are real questions from the question backlog and had been put in a recommendation table as can bee seen above in subsection "Recommendation logic".

After answering the questions, the participants are asked to chose the most fitting tipp themselves with regard to their partner. This is done without showing/communicating the by the recommendation logic chosen tipp. This enables a recommendation concept validation as can bee seen in Sprint 3 – Validate.

Furthermore, the participants are asked to add any additional tipps they seem fitting for their partner. In addition, participants asked in person by the team are asked to provide a free text tipp that would fit for their partner to even further fill up the tipp database.

At the end, the recommendation concept prototype shows the recommended tipp to the participants and the questionnaire ends. The results of the recommendation concept prototype validation can be seen in section Sprint 3 – Validate