Difference between revisions of "WinterMilestone 2 Ganler=edit"

From crowdresearch
Jump to: navigation, search
Line 1: Line 1:
'''Soylent : A word Processor with a crowd inside'''
+
'''Soylent'''  
  
Need
 
  
1. Human contributions into the interfaces
+
'''Requesters:'''
  
2. To shorten documents, proofread documents and edit parts of their document on demand.  
+
1. Make grammatical mistakes while writing
  
3. to correct Word, grammar spelling mistakes
+
2. Find it huge to correct a tense change
  
4. the task is huge to execute (proofreading)
+
3. Find it difficult to trim the size of the written paragraph
  
5. Trimming through paragraphs to fit in the size of the document.
+
4. Find it difficult to explain the task to workers
  
6. There is a need more than good interfaces aid and good artificial intelligences
+
5. Commonly make spelling mistakes
  
7. Need for easy word processing
+
6. Take help from friends and people they know to proof read or reduce the article size
  
8. Difficulty in writing – grammar, spelling mistakes.(Crowdproof)
+
7. Expects further help in complex tasks
  
9. Easy work to change tenses of sentences
+
8. Expects text correction delay time to be less – prefers real time
  
10. Large tasks for a writer for change in a decision
+
9. Is concerned about the task and content privacy
  
11. Help in complex cognition and manipulation tasks
+
10. Wants an interface which is understandable and helps him progress
  
12. Help from friends and other people
+
11. Waits for the correction to be clearly differentiable from the original data
  
13. Need to take care of these complex tasks for writers
+
12. Looks out for what part of the sentence has been compressed and which part has been expanded
  
14. Need to monitor spell checks
+
13. Looks out for people who can complete the task on time – get things in limited pages and within the deadline
  
15. Need to correct sentences without change in the sentence meaning (Shortn)
+
14. Is concerned and wants to confirm if the worker has understood the task – wants to confirm
  
16. Need for these text corrections to be done in real time
+
15. Wants the control in his hands to accept the corrections or not
  
17. Need for the correctors to understand the context
+
16. Wants to notify only the change required in certain pages or paragraphs
  
18. Solutions must maintain privacy, give timely output and be cost-effective
+
17. Finds it difficult to express the correction method correctly to the computer in case of scripting.
  
19. Need for the provided corrections themselves to be proof-read
+
18. Finds it difficult sometimes even to express to workers on what correction is to be done
  
20. Need for prevention of errors into the document
+
19. Wants to decide on how many people should work on his content
  
21. Need for the correctors to choose their comfortable area of correction
+
20. Wants to avoid lazy turkers waste his time
  
22. Need for the corrections to be notified to the owner
+
21. Similarly wants to avoid Eager Beavers as they complicate the task for him
  
23. Need for the writer to be familiar with the new correction interface
+
22. Wants to keep a check on if the workers are motivated enough while working
  
24. Need for the corrections to be done as quickly as possible
+
'''Workers:'''
  
25. Need to find the error, fix it and verify the error (followed by Soylent)
+
1. Find it easy to make corrections in already written sentences
  
26. Need for sentence compression
+
2. Generally help in article corrections
  
27. To maintain the content length within the time frame
+
3. Perform spell checks
  
'''Shortn'''
+
4. Have a cognitive load to maintain the sentence meaning right.
  
28. Need for the writer to express his expectation out of paragraph shortening
+
5. Find it not easy to every time get the article they understand
  
29. Need for the writer to provide real time feedback which correction takes place
+
6. Maintain the privacy of the content and task
  
30. Need for user to have control over the corrections (incorporated in '''Crowdproof''')
+
7. Sometimes unsure if they made the right corrections – expecting another proof-read
  
31. The writer should have the freedom to choose the paragraph or sentence which should be edited or corrected
+
8. Look for an option to filter their area of comfort in article selection
  
32. Need to efficiently convey the necessary methods of correction to the computer using scripting
+
9. Is concerned if the requester is fine with his understanding of the task – wants to confirm
  
33. Need to convey the necessary correction to be done efficiently to a human corrector
+
10. Expects the requestor to accept the rightly done work.
  
34. An intuitive interface for the writer and the corrector to understand what is going on  
+
11. Finds it difficult to understand the current interfaces on how to intimate the requestor on the changes made
  
35. The control in the hands of the writer to allow how many people need to work on his content
 
  
36. Need for the writer to correct his instructions as he gets misunderstood work from the turkers(correctors)
 
  
37. Need to verify of the correctors have understood the instructions or are they just working without even reading/understanding the instructions.
+
'''A Plea to Amazon: Fix Mechanical Turk!'''
  
38. To minimize lazy turkers from undedicated working
+
'''Requesters:'''
  
39. Avoid overwork to further complicate the task for the writers (Eager Beavers)
+
1. Find it much easier to navigate through amazon.com than through Turk  
 
+
40. Need to motivate the turkers to work efficiently.  '''(Find-Fix-Verify)'''
+
 
+
41. Need to generalize the number of turkers allowed to work on a task
+
 
+
42. Need for artitificial intelligence support to help turkers work better
+
 
+
43. Need for a standard communication pattern from the turkers side
+
 
+
44. Need for turker to be familiar with the topic to which he is about to contribute.
+
 
+
 
+
 
+
 
+
'''A Plea to Amazon: Fix Mechanical Turk!'''
+
  
Need
+
2. Frustrated on repeating the same thing again and again
  
1. to get the basics of crowdsourcing right
+
3. People are building on top of MTurk to improve their ease of work
  
2. to understand the foundation of MTurk and build over it
+
4. People find it very difficult and so are trying to make a better marketplace than MTurk
  
3. to scale up  
+
5. People want to Scale up, manage the complex API, manage execution time and ensure quality
  
4. To manage the complex API
+
6. Posting tasks should be easy for requestors
  
5. To manage execution time
+
7. They find the interface very difficult to manage
  
6. To ensure quality
+
8. Sometimes requestors are hiring full time developers to get the complex tasks done
  
7. Of a system which is more than just matching requesters and workers
+
9. It is very difficult for small requestors to grow
  
'''8. A better interface to post tasks'''
+
10. Very few requestors have one-pass tasks. And the requestors who have huge number of tasks are not comfortable using MTurk.
  
9. A easy way to search tasks and start work faster
+
11. Some users build and inframe-powered HIT misuse the system and get away
  
10. A better worker scrutinizing system for the requester to rely upon and majorly focus on providing tasks
+
12. For a long term user, this kind of additional personalized interface works, but not for short term users
  
11. A better interface which helps requesters easily manage work with a crowd
+
13. Requestors cannot easily differentiate good workers from bad workers
  
'''12. A worker reputation system'''
+
14. Getting the task done by multiple uninterested workers and get the quality check done is a huge frustration and a waste of time for the user
  
13. A system which helps the requestors differentiate good workers from the bad
+
15. User wants to get these questions verified
  
14. A system to retain the good workers in the marker
+
a. Does the worker have the proper English writing skills?
  
15. To have a more public qualification tests
+
b. Can the worker proofread?
  
16. To keep track of working history 
+
16. Want to provide rating without cognitive load.
  
17. To allow rating of workers
+
17. Requestor is the most affected guy in MTurk unlike the sellers in amazon.com
  
18. To disconnect payment from rating
+
18. Rating must be given from the side which is getting affected.
  
19. To differentiate spammers from honest workers who have produced different results
+
19. Requesters might have an idea or not have an idea of the time required to complete the task
  
20. To separate HITS and ratings by type
+
'''Workers:'''
  
21. To make all the mentioned points accessible from an API, which helps in automatic hiring decisions
+
1. Good workers are unable to get to the requestors
  
'''22. A requester trustworthiness guarantee for a worker'''
+
2. Wants to be rated correctly by the requestor
  
23. To show speed of payment
+
3. Wants to be able to appeal to the requestor on the work being rejected
  
24. To show the rejection rate for the requester
+
4. Want to be able to differentiate the type of tasks they are about to work
  
25. To show the appeal rate for the requester
+
5. Workers are not expert at all tasks, so they want to choose tasks only which they are comfortable at
  
26. To disallow the ability to reject work that is not spam
+
6. Workers want to rate the requestors according to their purpose.
  
27. To show total volume of posted work  
+
7. Workers wait for the requestor to start paying so that they can work further and rely upon the previous experience.
  
28. To make all the mentioned points accessible from an API
+
8. Check for the speed of payment from the requester
  
'''29. A better task search interface'''
+
9. Check for the rejection rate for the requester
  
30. A better way to predict the task completion time
+
10. Want to appeal for a rejection
  
31. A browsing system
+
11. Check the previous work ratings and experiences with the requesters
  
32. To improve the search engine
+
12. Like to know an estimate of how long it will take to complete the task
  
33. To use a recommender system to propose HITs to workers
+
13. Look for most recent HIT groups or the most HIT groups ignoring the smaller ones

Revision as of 00:20, 24 January 2016

Soylent


Requesters:

1. Make grammatical mistakes while writing

2. Find it huge to correct a tense change

3. Find it difficult to trim the size of the written paragraph

4. Find it difficult to explain the task to workers

5. Commonly make spelling mistakes

6. Take help from friends and people they know to proof read or reduce the article size

7. Expects further help in complex tasks

8. Expects text correction delay time to be less – prefers real time

9. Is concerned about the task and content privacy

10. Wants an interface which is understandable and helps him progress

11. Waits for the correction to be clearly differentiable from the original data

12. Looks out for what part of the sentence has been compressed and which part has been expanded

13. Looks out for people who can complete the task on time – get things in limited pages and within the deadline

14. Is concerned and wants to confirm if the worker has understood the task – wants to confirm

15. Wants the control in his hands to accept the corrections or not

16. Wants to notify only the change required in certain pages or paragraphs

17. Finds it difficult to express the correction method correctly to the computer in case of scripting.

18. Finds it difficult sometimes even to express to workers on what correction is to be done

19. Wants to decide on how many people should work on his content

20. Wants to avoid lazy turkers waste his time

21. Similarly wants to avoid Eager Beavers as they complicate the task for him

22. Wants to keep a check on if the workers are motivated enough while working

Workers:

1. Find it easy to make corrections in already written sentences

2. Generally help in article corrections

3. Perform spell checks

4. Have a cognitive load to maintain the sentence meaning right.

5. Find it not easy to every time get the article they understand

6. Maintain the privacy of the content and task

7. Sometimes unsure if they made the right corrections – expecting another proof-read

8. Look for an option to filter their area of comfort in article selection

9. Is concerned if the requester is fine with his understanding of the task – wants to confirm

10. Expects the requestor to accept the rightly done work.

11. Finds it difficult to understand the current interfaces on how to intimate the requestor on the changes made


A Plea to Amazon: Fix Mechanical Turk!

Requesters:

1. Find it much easier to navigate through amazon.com than through Turk

2. Frustrated on repeating the same thing again and again

3. People are building on top of MTurk to improve their ease of work

4. People find it very difficult and so are trying to make a better marketplace than MTurk

5. People want to Scale up, manage the complex API, manage execution time and ensure quality

6. Posting tasks should be easy for requestors

7. They find the interface very difficult to manage

8. Sometimes requestors are hiring full time developers to get the complex tasks done

9. It is very difficult for small requestors to grow

10. Very few requestors have one-pass tasks. And the requestors who have huge number of tasks are not comfortable using MTurk.

11. Some users build and inframe-powered HIT misuse the system and get away

12. For a long term user, this kind of additional personalized interface works, but not for short term users

13. Requestors cannot easily differentiate good workers from bad workers

14. Getting the task done by multiple uninterested workers and get the quality check done is a huge frustration and a waste of time for the user

15. User wants to get these questions verified

a. Does the worker have the proper English writing skills?

b. Can the worker proofread?

16. Want to provide rating without cognitive load.

17. Requestor is the most affected guy in MTurk unlike the sellers in amazon.com

18. Rating must be given from the side which is getting affected.

19. Requesters might have an idea or not have an idea of the time required to complete the task

Workers:

1. Good workers are unable to get to the requestors

2. Wants to be rated correctly by the requestor

3. Wants to be able to appeal to the requestor on the work being rejected

4. Want to be able to differentiate the type of tasks they are about to work

5. Workers are not expert at all tasks, so they want to choose tasks only which they are comfortable at

6. Workers want to rate the requestors according to their purpose.

7. Workers wait for the requestor to start paying so that they can work further and rely upon the previous experience.

8. Check for the speed of payment from the requester

9. Check for the rejection rate for the requester

10. Want to appeal for a rejection

11. Check the previous work ratings and experiences with the requesters

12. Like to know an estimate of how long it will take to complete the task

13. Look for most recent HIT groups or the most HIT groups ignoring the smaller ones