Skip to content

Latest commit

 

History

History
126 lines (108 loc) · 6.93 KB

Milestone1.md

File metadata and controls

126 lines (108 loc) · 6.93 KB

Milestone 1 : Ideation

Team Name: Random Member : Yujin Chung 20150691 Sangho Kim 20150148 Jaeryoung ka 20150007

1. Problem Statement

What is the problem your team is trying to solve?

Many CS / pre-CS students have troubles in making own academic timetable, specialized for unique circumstances they are confronting.

2. Problem background

How do you know this problem exists? Why is this problem important? Include both (1) external references (e.g., academic papers, news articles, or published surveys), as well as (2) internal investigation (e.g., results from making observations, personal experiences, or interviews with target requesters or workers). (maximum one paragraph)

​ Our team found this problem with the experience that all members saw at least one question about timetable revealed in SNS for every semester. As we searched from KAIST community in facebook 'KaDaejeon' to 'Kakaotalk' group chat, there are lots of people asking about the ‘load’ of their schedule. We also searched CS CA site, and saw lots of students want solution to their timetable-based problem.

img

[Figure 1] People asking for certain course's feature (e.g . prerequisite, workload), recommended courses based on own situation, the workload of a set of courses, course techtree recommendation in KaDaejeon

img

[Figure 2] According to CA(Counseling TA), there are ~2k counseling per year concerning with academic timetable problems.

3. Motivation

Why use social computing for the problem? Why not use machines or individual experts? (maximum one paragraph)

​ Not all courses and individual circumstances are well-known to an individual expert or specified by one trained machine-learning solution, because student's differs in courses. One more problem is that the features like workloads and interest are subjective so differ in various people. We also do not know the special circumstances a people confronts with such as military services. Some subjective questions can’t be explained by ML such as workload of that semester. Whether even the same timetable is affordable or not can be more meaningfully measured by the power of crowds, than by an algorithm from imperfect datasets. Like ox weight measuring experiments, gathering many people’s wisdom can measure these in a more meaningful and helpful way.

4. HMW Questions

How Might We (HMW) Questions: For the identified problem, discuss with your teammates what specific challenges exist. State these challenges as "How might we..." (HMW) questions.

  1. How might we make users easy to access and post questions and answers?

  2. How might we make the quality of the answer as high as possible?

  3. How might we make people ask earnestly?

  4. How might we make users get proper guideline for making own timetable?

  5. How might we filter unrelated posts for relevance?

  6. How might we enable short feedback time?

  7. How might we gain security for data?

  8. How might we make timetable upload easier?

  9. How might we make match asking the user (question) with the related user?

  10. How might we make people easy to see the proper already-existing questions and answers?

5. Solution Ideas

Solution ideas are your attempts at solving the HMW questions. Report at least 10 solution ideas per each of the top 3 HMW questions

  1. How might we make users easy to access and post questions and answers? (HWM 1)
    1. Make users easy to evaluate the timetable
    2. Gather initial users and q/a
    3. Support uploading image
    4. Make easy to access a site like mobile supporting, auto-login
    5. Make people connect at a high rate
    6. Anonymity available for users
    7. Make import timetable from "OTL+" images
    8. Make autosave function
    9. Extension with other social network services like Facebook, Twitter, and Instagram
    10. Pre-defined forms for posting
    11. Provide a “Quick Answer” tag so that one can get advice about smaller concerns in a faster way.
  2. How might we make the quality of the answer as high as possible? (HWM 2)
    1. The reporting system for trollers
    2. Blind system for trollers
    3. A system to judge the good answers
    4. Match questions with more-likely-to-be-related users
    5. Categorize the intention
    6. Get credibility values, such as 'reply from A, who took this course in 2019F'
    7. Reputation system for users who answer
    8. KAIST authentication system/ maybe get major values
    9. Sort of answers using quality
    10. Give a guideline and offer pre-defined forms for answering
    11. Make it easy to refer to another timetables or answers so that readers with some special concern can get more detailed information if they need it

  3. How might we make people easy to find and understand the proper already-existing questions and answers? (HWM 10)

    1. Categorize the questions and answers. Using Tags would be great, too.
    2. WRT 3-1, make similarity critique for each question
    3. Searching function for questions
    4. Giving evaluation function for each ‘timetable’
    5. WRT 3-3, show the evaluated result of the timetable in a visually easily-readable manner, rather than just showing numerical data only. (E.g. visually show that 72% of CS students thought that it is “hard”.)
    6. Premade the format so that answers can follow. This would make us analyze and provide the result in a more understandable way
    7. WRT 3-2, Autolink to similar questions
    8. User can answer with links to other question(related questions / already answered questions)
    9. Push alarms for my questions
    10. User can access to own questions
    11. Subscription function for questions (for user waiting for similar questions)
    12. Visualization techniques for answers, giving an overview of result for certain question

6. Storyboards

*Create 3 storyboard sketches, one for each of the top 3 solution ideas.

  1. Solution 1 - 4

img

  1. Solution 2 - 2

img

  1. Solution 3 - 12

img