Winter Milestone 6 BITS

From crowdresearch
Jump to: navigation, search

Introduction

This week we are collecting all the previous insights and developing a better system for creating a better taskfeed . Our system aims to create a good match between workers and requesters and providing an ordered Task list based on certain influencing factors which are:


1) SKILLS

The task list for each worker must be based on his specific skills for which he/she has been endorsed.

2) TIME

The task list must be ordered according to the estimated hourly rate of the workers.


3) REQUESTER

The tasks of a requester whom a worker has rated highly must be ranked higher in his task list.

4) Wages

The tasks must be ordered according to the minimum wage set by the requester , so as to filter those tasks which provide low wages.


These factors play a major role in worker's decision towards accepting a task and providing a better work quality. Workers want good and descriptive tasks which match their skills and hourly rate. Workers also need tasks which provide them a fair amount of hourly wage .


Modules

Module 1 :Task Tagging

Advantages

1) Tasks need to be tagged with particular skills , reputation scores and categories to uniquely identify the cluster of attributes required for the task completion , which would help in identifying those workers which match the particular workers which are capable of performing those tasks.

2) Task tagging helps in providing a short summary for the task , leading to better task clarity.


Disadvantages

Tasks tagged with specific skills might repel novice workers from the tasks even though the tasks might be fairly easy to complete. For Example-> A task requires programming , and the only thing required is as easy as writing a printf command , this can be easily learned by the novice workers but due to the tags they might simply avoid working on those tasks.

This can be better understood here

http://crowdresearch.stanford.edu/w/index.php?title=WinterMilestone_4_Adding_Tags_To_Tasks_For_Better_Categorization_And_Ranking


Module 2: Skill Endorsement

Advantages

1) Endorsement of workers by another requesters and workers results in honest tagging of workers , this would help daemo to find a good permuted task feed for each worker.

2) Helps in developing a better community of workers which endorse other workers honestly and accurately.

Disadvantages

Endorsement of workers with false skills and attributes may provide a false task feed.

This problem can be resolved by taking into consideration endorsement scores of the workers which would then help in creating a boomerang effect as explained here:

http://crowdresearch.stanford.edu/w/index.php?title=WinterMilestone_5_BITS


Module 3: Hourly Rate

Advantages

1) Estimating the hourly rate of the worker helps in selecting those tasks for the worker that matches his hourly rate.

2) Understanding his/her performance rate , the worker can work towards improving his hourly rate.

Disadvantages

The hourly rates provide an approximate guess for the hourly rate of the worker , this can lead to certain issues if the hourly rates are not properly guessed. A less hourly rate doesnt guarantee the accuracy of the work created by the worker. This should be


Module 4:Rejection Rate

Advantages

1) Tracking and measurement of rejection rate can properly help the worker to identify , if the worker wants to uptake the tasks given by a requester with a given rejection rating.

2) Workers need wages , and working on those tasks whose requester's rejection rating is high might result in wastage of time for a novice worker , and may lead to disappointment and bad rating for the requester.


Disadvantages

A high rejection rate might repel most potential workers due to the fear of not earning wages at the cost of their efforts, also a low rejection rate may attract bad quality workers which might result in bad work quality


Module 5: Good quality Snippets

For the betterment of workers and improving their skills , requesters can guide workers by providing good quality snippets for each task category , this would help confused but potential workers in creating work of good quality. The main aim of the crowdsource-platform must be not only to allow good workers to earn , it must also help inefficient workers in rising up the ladder and learning how to create good quality work .

It is very essential that the platform helps the workers is some way , proving them incentives on how to provide good quality work and how to select tasks efficiently , this would lead to platform's growth and also motivate and help inefficient workers to keep learning and create good quality work.

We do not see any disadvantages in this approach , as this approach only helps workers in becoming better , by the extension of the guild system , workers can also discuss better techniques on providing good quality work and reducing their hourly rate.


Contributors

@shivangi and @shivam