Vitality Platform
Data Driven Business Lab
Client company:Municipality Eindhoven
Gergana Agorastva
Remco Bisschops
Femke Boogerd
Nikita Gavrilov
Yaniek Martens
Lieke Nijs
Victor Plesciuc
Ilya Tsakunov
Aleksandr Vereshchagin
Project description
The main research question for this project was:
How can municipality Eindhoven create a platform with integrated intern and extern data for citizens of Eindhoven to find and fulfil their sport needs with matchmaking?
To answer this question we used the following sub questions:
- What are the end users’ requirements for a sport platform?
- How can sport activities that exist on different websites be combined?
- What is the most suitable matchmaking technique to show user’s sport preferences?
- How can the chosen matchmaking technique be implemented to display the user’s sport preferences?
- How can the matchmaking be ethically responsible and transparent?
Context
In this day and age citizens expect increasingly more transparency, accessibility and responsive services from their government (Gemeente Eindhoven). This can only happen by responding to citizens' wants and needs. The municipality of Eindhoven wants to create a platform that makes it as easy as possible to find the sport & facility activities that match the citizens' specific preferences. The platform will be used for all citizens of the city. To get the most out of the platform, there has to be some research on the current mismatches. Therefore, did the municipality created target groups to start customer journeys with. In this project can’t all customer journeys for all the target groups be looked at. Besides serving every citizen in the city, there also are two defined specific target groups where in this project can take a deep dive into their customer journey. These defined target groups are expats and youth from 12 years.
Municipality Eindhoven decided to choose expats/internationals as a target group because they are very much skilled to use digital tools but are not familiar with the Dutch sports system. A tool that gives them quick and easy insight into their sport needs, could therefore be very impactful for them. The government hopes people will do sports during their entire lifetime. The municipality does know that doing sports at a young age improves the chances that people will do sports when they’re adults. Therefore, the second target group is the youth. This project will focus on children of 12 years and older, because it turns out that many children stop doing sports in this category.
Results
The result of this project leads to the proof of concept of the first step of the initial municipality’s project plan. In the past six month the Datastic group has researched, developed, test and validated the initial concepts that were advised by Bureau Moeilijke Dingen such as obtaining information from web using webscraping techniques, matchmaking possibilities, law and ethics investigation. In this chapter the results of above concepts will be discussed as well as other Portfolio assets of Datastic.
Portfolio assets:
- Project Plan
- Research Document
- WebScraping prototype
- Matchmaking prototype
- Law and Ethics investigation
The municipality asked Datastic to create a prototype based on the webscraping and matchmaking part of the platform of vitality. Datastic divided this project in separate phases where each phase had their own outcome.
Firstly, during the first phase end-users liked to see the description, timetable, place, location, recent activity, age, experience level and price for each activity within the application. In the second phase of the project, the web scraping took place. During this phase Datastic figured out that almost all of the websites didn’t contain open activities. There is no content to scrape and the HTML structure for each website diverse a lot compared to others. Also, the content that end-users wanted to see in the application, is not available at the activities. Datastic advice to do one of the following things:
Create your own websites where associations or commercial companies can put their activity into a form. This means every activity will be putted into the Municipality’s site in the same structure, so it is easy usable for their platform (this will reduce the costs) and every needed content can be asked to the organization.
Contact the organizations about your plans and perform research if they are willing to update their websites with the content the municipality likes to scrape.
In the third phase matchmaking took place. There were a few options available to do the matchmaking with, like machine learning and a rule engine. Datastic chose for a rule engine because the machine learning wasn’t possible with the available information. Datastic advice the municipality to keep using the rule engine at the first phase of the platform. This rule engine fits perfectly to the desired swiping solution the municipality wants. In a later stage of the platform machine learning could be added by the municipality which can offer activities to the users based on their previous preferences. The municipality must consider which solution they will use because of the complexity of the model.
During the phases also the ethical part was taken in mind by Datastic. Based on different core values a couple of target groups were determined. These target groups all have different expatiations about the application. Datastic advice the municipality to use these expectations when they are developing the platform. Also take in mind to look further then the used core values.
Methodology
What are the end users’ requirements for a sport platform?
Strategy: Field and Workshop
Interview: Interviews held with people in the focus group
Focus group: Conclusion from interview is discussed with the focus group
Requirement prioritization: With the conclusion from the interviews and discussions a list of requirements is made using the MoSCow method
How can sport activities that exist on Eindhoven Sport website be combined?
Strategy: Library and Workshop
Literature study: Datastic did research on languages and platform that can be used for webscraping.
Best, good & bad practices: Datastic did research on the best way to webscrap and what common mistakes are made with websraping.
Expert Interview: Held expert interviews with the people from the work field who have a better understanding about webscraping.
Prototyping: A prototype was made to webscrape the website’s that the municipality had choicen, with conclusion that most websites are unusable.
What is the most suitable matchmaking technique to show user’s sport preferences?
Strategy: Library, Steppingstone and Lab
Literature study: Datastic did research on machine learning and rule engines. For machine learning there was also looked at clustering, K means algorithm and decision threes.
Best, good & bad practices: Datastic did research on the best way to make a rule engine and what common mistakes are made with making a rule enigne.
Data analysis: Because no information could be webscraped, a file with mock up data was made.
Persona: Datastic designed a persona to create reliable and realistic representations of the key audience segments for reference.
How can the chosen matchmaking technique be implemented to display the user’s sport preferences?
Strategy: Workshop
Prototyping: Datastic made a rule engine using a score system.
Ideation: Future steps that need to be made are data storage and collaborative filtering. This is to start using machine learning.
How can the matchmaking be ethically responsible and transparent?
Strategy: Workshop and Showroom
Brainstorm: The ethical responsibility from the vitality platform is discussed with the help of a counselling ethics discussion with use of a brainstorm.
Product review: With the TICT Tool there was looked at the value Inclusivity, which was one of the values that was found to be important in the brainstorm session.
Requirement prioritization: With the brainstom session and the TICT Tool it emerged the top three values are important for this project.
About the project group
Our group exists of people from different backgrounds. 4 project members have an ICT & Business background, 3 project members have an Industrial Engineering background, 1 project member has a Finance background and another project member has a ICT & Software background. 5 project members are international students while the rest are Dutch.
In our project we always had a meeting on Monday to discuss the plans for the coming week, and on Friday we would have a meeting to show our results with the rest of the company. Most of the time on Wednesday we would have meetings with the Municipality of Eindhoven to discuss our process of that week.