Smart Sticky!

Creating and sharing ideas using Smart Sticky!
Log on to the workspace
Ideas grouped by clustering functionality of Smart Sticky!
Posting an idea
Relevancy Graph: visualizes the relationship between groups of ideas
Timer funciton: helps your team to create ideas during a limited time
Timer funciton: dialog to tell you the end of time
Statistics function: numbers of posts of each members
Export frunction: you can download CSV file of all the sticky
official character

Inspiration

It is usual to brainstorm your idea using sticky when you come up with something creative idea. We often follow these steps below.

Step1. Record each idea on sticky.

Step2. Look for ideas that seem to be related.

Step3. Sort sticky into groups until all sticky have been used.

We all have to get together in the same place to do these, but it’s sometimes kind of difficult to be together. And it’s pain to divide all the idea we hatched into some groups.

To solve these problems, we built the app called “Smart Sticky!.”

What it does

Sharing your idea without getting together. You can put out ideas whenever and wherever you want if you have your iPhone or smartphone!! Share your creativity ideas with your company on the app! Save your time, think more.

Smart Sticky! can group all the ideas automatically. So you can spend your precious time for thinking something creative more!

Naming group title based on the feature of ideas. Because Smart Sticky! makes each group named, it helps to find the unknown field/ area of idea you never thought.

How we built it

-User Interface

Built this User Interface using Node.js, Vue.js and Vuetify.js

-Grouping Engine

Auto-categorizing of ideas Your idea is going to be categorized as soon as you post your idea based on the unsupervised learning model by the K-means algorism.

Re-learning by Re-grouped ideas This is composed of the supervised learning by Logistic regression algorism based on the training data using the result of the grouping by users.

Challenges we ran into

Three challenges we ran into are…

Machine Learning

This is our first challenge to use scikit-learn which is python’s machine learning library to implement the machine learning without using out-of-the-box API such as IBM Watson. Using the library, we implements clustering all the ideas on the app.

Modern Implementation for UI

We also tried to use Vue.js on Node.js. This is kind of popular, trend and modern way to implement, and we did.

Realtime Update

Using Firebase which is the realtime database, we achieves reflecting the latest data immediately.

Accomplishments that we're proud of

Architecture and User Interface

Accomplishment of Sharing all ideas online immediately. You can post your own idea wherever you are at. To achieve this architecture, we adapt the realtime database, FireBase.

Grouping Engine

The grouping engine we built combines both clustering and logistic regression. This sophisticated hybrid system would provide the user more intuitive grouping.

What we learned

We learned 3 things below.

The importance of “Let’s just give it a try!” Choosing the application development environment which suits for what you want to build. The importance of making choice of the tasks in the limited time.

What's next for Smart Sticky

Implementation of VR. In the virtual space, you all can see all the floating sticky and divide into other group using your finger. You can chat with your company in the space too. Moreover, you can choose the location where the sticky shows up like the Southern island. It helps you to imagine something fun!

Handwriting

Implementation of handwriting on sticky. It is said that handwriting makes your brain more active. So, by building the input interface of handwriting, Smart Sticky! will help to make user imagine more creative idea. We are happy to use the APIs of Tegaki produced by Cogent Labs.