top of page

EZcook

​Responsibility

Information Architect, Concept, Interaction, Visual, Test

Role

UX/UI Designer

Team

NTHU Computer Science Undergraduate Project Exhibition

Period

2017.9 - 2018.1

OVERVIEW
OVERVIEW

EZcook is a HCI study on how to enhance the efficiency of learning physical task. We took cooking as our research target and developed our design based on crowdsourcing and gamification mechanism. In the end we built a website that allows novice learners to start live streams and audiences to watch them and give instant advices.

INTRODUCTION

"利用社群外包讓業餘人士來給予新手學員建議可以降低教學成本,且遊戲化過程來提高即時參與度這是一個很好的想法"

--- From reviewers, TaiCHI 2017 Demo Session

EZcook is a live stream platform which can match the cooking novice learner and community who loves food, solving the problem of hard to learn cooking.

One of the biggest problem of learning cooking in traditional way, is hard to find someone to teach the learner side by side. Even if the learners find someone to teach them, they tend to follow the steps without thinking, which leads to learning nothing. We thus develop this platform, trying solve these problems.

First of all, cooking is a continuous process, every question should be answered immediately. So we came up with making whole process live, which means the learner start a live stream of cooking, and audience would give the learner advices. We also apply the concept of "gamification" and "community sourcing" to solve the problems. Community sourcing give us great amount of answers. Based on the former research, the more answers, the more correct the answers tend to be. To make audience willing to give their advices, we design a "gambling system." When a audience's advice is chosen useful by the learner, then he or she will get a credit, which can exchange rewards.

CHALLENGE
  • Learning skills using live stream is a new topic, regardless learning cooking. There's no reference and standard for us to review or follow, every design will face the difficulties of fitting behavior and mental model.

RESEARCH

Contextual Inquiry

n our project, we interviewed cooking experts and cooking novices separately. In the novice section, we invite a novice to describe her recent experience in cooking, with or without any difficulty? Or any inconvenience? For the experts, we asked him the latest experience of teaching others cooking is how? Is there any point bothering him?

We arranged the results of the interviews into affinity maps to find the users' needs. We found that learners value these themes:

  1. learn from communities

  2. customization of teaching

  3. knowing detail of preparing ingredients

  4. be answered questions during cooking

Persona

We built two personas, a receiver and a provider that show their different behaviors, goals, and motivations.

Hans

He is a junior and he studied electrical engineering at school. Even though the electrical engineering department is busy, he still surf Facebook from time to time. He is usually an active in the department, and he has also been a chief of activity in the EE camp.

He also participated in the cooking club, in which the seniors first explained the recipes and provided assistance when needed. However, he felt that most of the recipes need oven as the tool, and he would not have the chance to redo them once he returned home. In addition, the recipe steps are not very detailed, and many things are forgotten.

Occasionally he will watch Aki and Gordon Ramsay's video on youtube. When he sees a beautiful dish, he will want to make it. However, because he spent more than an hour making fried rice for the first time, and finally the rice was not cooked, and the taste was too salty, so that his motivation after that was greatly reduced when thinking about the possibility of failure. He thinks that if there is a product or service that can improve the efficiency of learning to cook, and reduce the sense of anxiety in the cooking process, it should be a great idea.

May

May is a 35-year-old office worker who is usually a very strong department head in the company. She likes to read articles related to business, psychology, and education. At the same time, she is also a senior Reddit user, and she likes to chat on Reddit to get away from the pressure of work. Occasionally, she will also surf Facebook and Instagram to see some entertainment news or friends' posts, posting some finished food photos.

 

Sometimes her friends will ask her in the comment that how to do the dishes, and she will share the recipes she finds. Or when she is asked for details, she will reply directly with a message or comment. Sometimes she would like to write some instruction on Facebook. However, due to the time is shredded, she has never put it into practice.

 

She and her husband are usually very busy, so although May would like to cook, they do not eat at home everyday, but they still like to watch the food or cook shows at rest. Compared to purely foreign cook show, she prefers Taiwanese show such as Chef Aki or Chef James, because foreign countries pay more attention to the beauty of chefs' cooking, while domestic ones emphasize the details in the process of cooking. However, the storytelling series like Hell's Kitchen will attract her, because this is not for learning how to cook, so that the details are not so important.

 

She thinks that if there is a product or service that can interact with other cooking groups and can create value through sharing experiences, it should be a good idea.

DESIGN

Brainstorming

We brainstormed and generated numerous ideas, and converged into 2 main ideas:

  • community sourcing via live stream

  • gamification mechanism 


​Since cooking is a continuous process, every question should be answered immediately. Live stream is a great way to get comments or feedbacks, and learners can interact with audiences simultaneously.

Challenges

  • How do we attract our audiences and ensure the quality of answers?

  • Also, in fact the live stream technology cannot really achieve the goal of interact immediately. How to solve this problem?

Functional Flow

To attract our audience to come in, we design our gamification mechanism. Also we hope to achieve immediate feedback through a gamification mechanism of gamification, and our gambling mechanism is as follows:

When watching the live broadcast, the audience observe what learners are missing, they first guess which mistake will take place,  and "bet" that in the way of providing a solution to the problem. The solution will be provided to the novice. If novice chooses their solutions, they won this gamble, and they can get rewards. We use this method to improve the delay of live broadcast technology and increase the audience's willingness to provide answers

Wireframe & High Fidelity Prototype

Landing Page
Recipe Step
Live Stream
Profile Liked
Live Room(Community)
Develop

We built a website as our MVP and used it for testing. For the front-end, we used html / css / javascript to build, and for the back-end we used firebase to record and extract the log when conducting the test.

EVALUATION

Recruitment

Since the challenge were come from the audience's side, so we test our audiences first. We designed a questionnaire to make sure we found the right testers. We listed out some attributes as the filter of the background, and we interviewed the professionals to define 7 levels of the experience of cooking.

​We recruited 4 testers with 1 female and 3 males. 2 of them has more experiences in cooking while the other have less.

Conclusion & Key Findings

Experiment Design

We started a live stream on facebook first and embedded it into our website. We would first give them the same instruction of using our website. Our website for the audience was divided into four section:

  • Live Video

  • Comment

  • Recipe

  • Suggestion & Bet

What they should do was to watch the live stream and interact with the cook and other audience, which were all played by us. Our system would record what they say and what they suggest, and also we would record their screen to see how they use and interact with our website.

Users generally liked to interact with the cook/learner(which was played by us), while there were several usability problem, our study pointed out interesting behaviors of audiences.

  • Since the pace of cooking process is very fast, audiences will simplify their comment / feedback / answer.

  • Audiences tend to interact in the betting section since the cook/learner notices the comment there more easily.

We dived into the users' log and did the affinity analysis to their chat.

The comment was divided into two groups including

  • The answer to the question from cook/learner,

  • The direct instruction for what to do next

If grouping the sentences based on the meaning, there are six types of them including

  • ​Question

  • Answer

  • Advice

  • Review / Rate

  • Assault / Insult

  • No meaning

© 2020 Portfolio by Daniel Lee

bottom of page