Carol Hsu's profile

WIKIMEDIA CH Data Science Challenge for openedu.ch

WIKIMEDIA CH Data Science Challenge for openedu.ch
Empowering Knowledge — An open source educational platform for educators
Project Summary
OPENEDU.CH is a platform that advocates for fair and equal education. To attract new users, we revised the platform's infrastructure (back-end and front-end) and introduced a new ontology for data management. A redesigned website prototype is provided to demonstrate new features, including data query and NLP (Semitic search) function.
About OPENEDU.CH
OPENEDU.CH is an open education resource platform launched in 2020 by the WIKIMEDIA CH organization. It advocates fair education, and encourages educators to connect, collaborate and share knowledge through this platform. Through this platform, educators can benefit from accessing educational rescore on any topic, anywhere, in addition, OPENEDU.CH also provides training services and tools upon request. To achieve a business goal, and build such a platform, a strategic business analysis and data-driven approach are needed.
The Process
To achieve the business goal, the Double Diamond Process Model is applied as this model is not only focused on design elements but taking a holistic view to design a system that can be run technically and visually. The Double Diamond Process Model consists of four phases as follows:
Double Diamond Process Model
Discover:
 ● Understand what is ontology; who are the key stakeholders, and how they will be 
affected by the ontology.
  ● Reverse the existing database structure, and design, function, and service that offer on the current website

Define:
  ● Gathering insights from all team members and defining the domain and purpose of
ontology and formulating the structure of the database
  ● Identify key problems that need to be addressed now. we decided to focus on optimizing searching, filtering and uploading functions.

Develop: Once we define our key problems, and common goal, we divieded tasks based on member's expertise, which are Database, Data science and UX|UI, and each member can focus on developing solutions

Delivery: we proposed three actionable solutions which are implemented within six weeks
  ● Refine ontology, data model and data pipeline
  ● Search engine optimization with NLP approach
  ● Visual Prototype: redesign landing page, result search page and upload form
Discover

Assessment
The current website adapted a simply design, nevertheless it is unclear who are the main audience while for visitor to understand what service or products the website offers or what benefits visitor will get through the website. We first reversed the existing website, database and data modeling to identify initial problems for further adjustment.​​​​​​​
OPENEDU.CH landing page
Usability and User Experience Assessment
The three main categories that were evaluated were value proposition, usability, and user experience. Initial issues and corresponding improvements are described in the table below.
Competitive analysis

Open Education Resource(OER): Medium, Google Scholar, MIT, Oercommons, BC campus and Europeana are chosen for benchmarking due to the platform type, content, and service they offer are very alike to OPENEDU.CH. Based on these five main criteria: information architecture, navigation, search function, filters and search results, we can further refine the OPENEDU Website.
Competitive Analysis
Persona

In addition, we also create 2 personas to get a better understanding how what OPENEDU.CH's users probably look like and be able to come up with better solutions to tackle their problems
Persona: uploader and modifier
User Journey

A user journey includes needs recognition, exploration, evaluation, decision making and post behaviour. The diagram map out critical touchpoints, actions and emotion of the target personas
User journey mapping
User Interview

After completing the above analysis, key findings are extracted and formed five key subjects that we would like to further investigate through user interviews. The subjects are:
     1. The context: In what situation educators would visit an educational resource platform
     2. Behaviour: What do educators look for when visiting educational resource platforms
     3. Motivation: What is the motivation for using an educational resource platform
     4. Expectation: What product/ service do you expect to see on an educational resource platform
    5. User Engagement: What is the motivation for engaging with other users on an educational resource platform
These five questions are further elaborated into eight subquestions. To improve the usability and user experience of the current site, user testing was also given at the end of the interview.
User Interview: questionary design
We conducted user interviews with 5 potential users (one internal and 4 external users.

     ● 90 % of the Users search bar to start their search on the majority platform
     ● Online Community, Peer review and students feedback are critical touchpoints
     ● Having clear instructions and guidance would help the user to complete the uploading process
     ● Users wish to get in touch with other teachers and connect in a simple, easy way
Define​​​​​​​

We collected ideas from team brainstorming, user interview and competitive analysis, and we finally selected the most prominent features which are: Search UI, Filters, Uploading Form, and Redesign Landing page that is aligned with solutions from our technical team, and be able to visualize in the final prototype
Develop

At this stage, we compared potential solutions and selected the most reasonable ideas to develop and listed in the table below:
Features Development
User Flow

After we define the key features, I designed the user flow of the search and file uploading process.
User Flow: content search, registration and file uploading verification
Design: UI Redesign
In the design phase,  Mental Model and Design Principle is adopted as the model emphasis human's behaviour in the real world context.

A mental model is a sort of thinking system in our brains. It guides people to react and interact with something based on what people see and what they believe in their thinking system. Users expect what they believe and will interact with a product and service they have experienced before. Therefore, making the website similar to its competitor can help users easily navigate the site without hesitating.
Landing page

To eliminate overload information on the landing page, and to optimize users experiences and site usability, four features are redesigned which can be seen in the proposed landing page below:

   ● Core message: an engaging slogan tell users whowhat (service and products), how  to use the site
   ● Enlarge Search UI: Based on user interview, we found majority users are search dominated and thus making the search bar visible on the landing page can help user to quick find content they need
   ● CTA button: ask people to sign up and explore the site
   ● Visible navigation: make the navigation visible to users without opening the hamburger menu 
Resource page: Vertical Filter bar, search output and recommender features

Here I use three columns grids to divide filter, search result and tags features. As the article, feature is the most important and therefore placed in the middle. 

Vertical filter bar: replaced the original horizontal filter. The vertical filter is composed a list of dropdown menus, and each menu is ordered by its lexicon hierarchy, which can help the user to identify the category they wish to use for filtering content.  

Article card: displays information of article. Primary information such as author/ institution, title and subject line are displayed with heavy and bold fonts. While secondary text filed such as short description, label (project type, media type, format) and update date are displays in small or light fonts. 

Recommender features: trending and recommend topics are presented with label UIs 
Resource page
Sign up pages

On the sign-up page, we want to collect more user basic information, not just their name and email. However, to make the user feel comfortable and share their information with us, we implemented some tags and a drop-down menu to help users to fill out their information in a simple way. 
Uploading Pages

We make the uploading form simpler and divide the process into four steps, to make the user informed and know which section they are and be able to estimate how much time they need to complete this task.
Result/ Delivery: Interactive Website Prototype

A landing page and resouce page ared created in Figma. In addition, three interactive features can be test in this website prototype:

Search bar: try to entry "scie" and select science competition then press enter to see the output search result

Filters: a list of dropdown menu allow users to interact with

Sign up and uploading form: you can access sign up page by click the sign up button, it will guide you through the sign up and add file processes
Go to the Market

Now that we had achieved the outcome. our team also design and pitch desk to support the launch of these exciting features that we have developed over the past six week.
Reflection

The purpose for this project is to provide tangible and actionable solutions for the client, although the redesign website was not required by the people, we believe having a website prototype that can visualize our solution in a feasible way is more convincing when pitching our idea, There were some flaws in the process and improvement that need to be addressed accordingly based on the data collection in the future development.
WIKIMEDIA CH Data Science Challenge for openedu.ch
Published:

Owner

WIKIMEDIA CH Data Science Challenge for openedu.ch

Published:

Creative Fields