CS492: Crowdsourcing

Project Gallery: Voluntary Crowdsourcing Applications

Students worked on a semester-long design project, in which they designed and built a crowdsourcing application that involves a VOLUNTARY OR INTRINSICALLY MOTIVATED CROWD. This poses interesting and unique design challenges in terms of task design, incentive design, and quality control.

Seven student teams came up with creative designs, and this page gives a snippet of their design process.

Let's Read

Members: Junsoo Park, Youngbo Shim, Sang-Gyun An

Researchers usually face difficulties in reading scientific papers of unfamiliar fields with a good level of understanding. We introduce a co-reading platform ‘Let’sRead’ where readers naturally share their reading traces which are aggregated into a helpful visual guide for efficient reading flow. Different from any other highlighting tools, it dynamically update the status of highlights according to weights made in real time by readers.

Check out the LIVE INTERFACE

  1. Problem Identification and Ideas
  2. Tasks, Sketches, and Video
  3. Pitch
  4. Low-fi Prototype
  5. Mid-fi Prototype
  6. Interactive Prototype
  7. Final Presentation
  8. Paper

LetStory

Members: Miroslav Masat, Oscar Johanson, Goh Wei Xiang, Liu Ching

LetStory is a crowdsourced content creation tool promoting writing with a cause. In this application we demonstrate our concept of collect/create/vote, that produces diverse and media rich stories from crowd.

Check out the LIVE INTERFACE

  1. Problem Identification and Ideas
  2. Tasks, Sketches, and Video
  3. Pitch
  4. Low-fi Prototype
  5. Mid-fi Prototype
  6. Interactive Prototype
  7. Final Presentation
  8. Paper

Livelight

Members: Yungi Jang, Jiwoo Park, Muhammad Umair

Generating highlights for live videos such as sports match, presidential debates etc. needs experts and also viewers might have to wait for some time in order to watch those highlights. We present LiveLight which uses every viewer's individual moments (moments they love) to generate highlights during the live stream. As far as we know, our proposed system is first of its kind that uses crowdsourcing for highlight generation. The unique feature of our system is that it processes crowd's moments during the live stream and generates highlights before the live stream ends.

Check out the LIVE INTERFACE

  1. Problem Identification and Ideas
  2. Tasks, Sketches, and Video
  3. Pitch
  4. Low-fi Prototype
  5. Mid-fi Prototype
  6. Interactive Prototype
  7. Final Presentation
  8. Paper

1DNQ

Members: Sunggeun Ahn, Young-Min Baek, Sungjae Hong

There is a call for effective environment for learning and education. As a solution, team Meddler proposes 1-Day-N-Questions (1DNQ) platform, which provides crowdsourced fresh questions number of times a day with alarm buzz. What makes 1DNQ different from other wordlist-based solutions is the presence of a study-group-like clustering system, which enables to motivate learners through poking other members by making a question.

Check out the LIVE INTERFACE

  1. Problem Identification and Ideas
  2. Tasks, Sketches, and Video
  3. Pitch
  4. Low-fi Prototype
  5. Mid-fi Prototype
  6. Interactive Prototype
  7. Final Presentation
  8. Paper

Minutu

Members: Yekaterina Abileva, Nurzhan Yergozha, Noé Domínguez

Provides students with the ability to leave a fine-grained feedback on the lecture and specific parts of the lecture linking it with the emotion they experience about it. From the other side, provide professors with the ability to see all the feedbacks and design their lectures according to it. They can also review some confusing concepts from the previous lectures or give more information about the class interest. The key concept which is different from similar systems is connecting a feedback with a specific emoji.

Check out the LIVE INTERFACE

  1. Problem Identification and Ideas
  2. Tasks, Sketches, and Video
  3. Pitch
  4. Low-fi Prototype
  5. Mid-fi Prototype
  6. Interactive Prototype
  7. Final Presentation
  8. Paper

PAPERGALAXY

Members: Hoon Kim, Hyun Sung Cho, Juho Sun

Understanding how each research paper is connected to each other is very important for researchers and students for them guide themselves in the wide world of science . PAPERGALAXY aims to create a semantic network of academic papers that represents how papers are related to one another. Unlike other solutions, PAPERGALAXY uses crowdsourcing to find connections between paper, in which every connection between researches is created by the crowd.

Check out the LIVE INTERFACE

  1. Problem Identification and Ideas
  2. Tasks, Sketches, and Video
  3. Pitch
  4. Low-fi Prototype
  5. Mid-fi Prototype
  6. Interactive Prototype
  7. Final Presentation
  8. Paper

Verbivore

Members: Oisin Daly Kiær, Paul Grau, Yoo Jin Lim

Verbivore is a vocabulary flashcard studying and -generation application that attempts to lessen the tedious busywork associated with flashcard creation, while also using the crowd-work of its users to iteratively improve its set of cards. Users study by themselves, while being confronted with microtasks that help them learn, while we use their tasks’ output to improve the cards themselves. Iterative improvement towards a set of ‘holy-grail’ perfect vocabulary flashcards.

Check out the LIVE INTERFACE

  1. Problem Identification and Ideas
  2. Tasks, Sketches, and Video
  3. Pitch
  4. Low-fi Prototype
  5. Mid-fi Prototype
  6. Interactive Prototype
  7. Final Presentation
  8. Paper