以評價標籤建置貼文推薦排序演算法
When we brainstorm to get some valuable idea, we always need a lot of time to discuss. Then we find the key idea after each step ends. But that’s a difficult work to search the key idea in the long historical record of our discussion.
So, need a system to classify each idea is promising or not.
PIs (Promising Ideas) are defined as those that sustain participants in design mode work and that advance knowledge building. Moreover, PIs are those that students wish to spend time on, that may change in further inquiry and that would deepen their shared understanding. (Chen)
So, we have five key points to classify promising idea:
- SP : Sustain participants in design mode work
- KB : Advance knowledge building
- WST : Students wish to spend time on
- CI : Change in further inquiry
- DU : Deepen their shared understanding
Then we have a prototype like this:
Each user could create the new root of the discussion or build on some idea to the existing chat node. Beside build on and create, the user can read other user’s build on, then highlight some text, giving a suitable tag. Users could highlight a text and give tag one or more time.
