GR1 - Viral content generator

Group Members

1. Oliver Song
2. Marco Alagna

Problem statement

Generating viral content for the internet is currently an inefficient process involving either photoshop or web services (www.memegenerator.net). Furthermore, they are restricted to "caption image" type memes. This is not representative of the pool of meme types actually on the internet. Finally, this method of meme generation is not mobile; this is particularly egregious because the large majority of meme source content comes from mobile devices.

We find this to be onix-eptable.

User Analysis

Our users come from a variety of backgrounds. Our most representative user demographic will have:

• spent at least 2 years understanding internet culture
• high school or college level education
• own a smartphone (iPhone primarily)
• either gender
• a sense of humor

Alternatively our target user population can be split into two groups: those who understand internet culture and those who do not.

Those who understand internet culture:

• will have an understanding of which expressions can go with which captions
• will have a quick and accurate judgement of what will and will not be funny to the general internet population
• will be best suited for a traditional meme maker as in "take picture, add caption"

Use case:

We interviewed a person in this space of user population; we'll call this person Subject A. Subject A is a student at Princeton in his second year. He has approximately 8 years of internet browsing experience. He states that he has "high proficiency" in determining web trends and popular memes, as well as a "proficient" meter for judging whether something will or will not be popular. Subject A likes to make images featuring his friends, with funny captions that they would say in certain situations. He expressed his frustration with the current state of content generation technologies. He stated that what he used (memegenerator.net) was slow, buggy, and featured many ads. He described his general usage scenario as:

1. Brainstorming
2. Search for picture of friend
3. Crop background out
4. Going to www.memegenerator.net
5. Click on upload photo
6. Attempt to upload right file size
7. Redo a couple times ("unknown file upload error")
8. Upload correctly, input titles
9. Wait for image to become available
10. Post on friend's wall

In total, this costed Subject A between 30 and 40 minutes. We believe that this can be done more efficiently.

Those who do not:

• will not understand what the general internet population would find funny
• will be best suited for nontraditional viral content generation, such as "video filtering" or "song making", which make use of simpler emotions that do not employ much of the "meta-humor" in many internet memes 

Use case:

We interviewed a person in this space of user population; we'll call this person Subject B. Subject B is a student at Brown University in her first year. She does not spend much time on the computer, and her exposure to viral content is mostly through shared posts on Facebook. She would say that she has 1 year of internet browsing experience, but she doesn't really know. She states that she has "no proficiency" in determining web trends and popular memes, as well as a "semi-proficient" meter for judging whether something will or will not be popular. Additionally, she states that she would not mind trying her hand at viral content generation.

Her ideal usage scenario is:

1. "Take a picture or something"
2. Label it
3. Post it

While Subject B did not really understand much about viral content, her model usage scenario is representative of a large population of users and we believe that our final procedure should be as close to her model as possible.

Task Analysis

Source of humor (photo, video, or sound, or any data)

 - methods of capturing data

Processing data

 - cropping

 - clipping

 - etc

Labeling/categorizing humor

 - captioning

 - creating gif

 - etc other labeling

Publishing

 - Social networks

 - Account management

 - Syncronization

 - Tagging

Viewing ( optional )

 - view your uploads in one feed online

1. Why is the task being done?
   1.  
2. What does the user need to know or have before doing the task?
   1.  
3. Where is the task performed?
   1.  
4. What is the environment like? Noisy, dirty, dangerous?
   1.  
5. How often is the task performed?
   1.  
6. What are its time or resource contraints?
   1.  
7. How is the task learned?
   1.  
8. What can go wrong (exceptions, errors, emergencies)?
   1.  
9. Who else is involved in the task?
   1.