Difference between revisions of "WinterMilestone 1 urumisena"
|Line 24:||Line 24:|
====Likes (thinking like a requester): ====
====Likes (thinking like a requester): ====
* Capabilities: Requesters can find capable people easily (they are not restricted by geography).
* Freelancing: Requesters don't want to hire people for their companies because is a small work.
* The experience of creating a survey: I think it can be improved because, if it's the first request for an employer or he/she just didn't understand the flow, it can frustrate him/her.
* Unknown time: My task wasn't resolved. For a requester, don't knowing the time it would take to get the job done it's stressful (maybe nobody'll take the task or the employes needs the work to be done ASAP).
Revision as of 06:24, 17 January 2016
- 1 Team urumisena
- 2 Experience the life of a Worker o UpWork (formerly oDesk)
- 3 Experience the life of a Requester on MTurk's Requester Sandbox
- 4 Readings
Experience the life of a Worker o UpWork (formerly oDesk)
I'm not living in USA then I used this platform. The connects thing immpressed me, I didn't know I needed something like coins for applying for a job.
- Skills evaluation: using this tool I could prove my skills (this can be useful for a requester too).
- Profile-based recommendations: The tool shows job opportunities related with the field I selected in the profile section.
- Profile creation: I spent more time thinking in which sector I can be located and what to write in the overviews than in reading, reflecting and writing about the whitepapers.
- This profile doesn't allow me to include all my strong points and it was limited to one field.
- When I finished the loong profile, "Hey! we don't know enough about you then you can't search for a job"... Then why everything was approved? I felt stressed and bored.
- The review of my profile took more time than I thought then I lost some task I was interested in. I would like the pltform telling me how many time I need to wait for being able to send proposals.
Experience the life of a Requester on MTurk's Requester Sandbox
I conducted a survey experiment on MTurk's Requester Sandbox because I'm not living in USA. For me it was hard to create the survey because the "interactive" field: I don't have a wizard or easy and understandable options for creating the questions.
I changed the text of the questions but, when I tried to change or delete some questions, all my work was lost! I felt uneasy and manipulated the HTML code but, if you're not familiar with it, you can feel lost. What if I want to create another different type of questions?
Finding the task was troublesome too... I filtered the tasks but this didn't worked so well and it was pretty hard to find my own post! I think for a requester could be helpful to have access to the "location" of his/her tasks because they can be sent and shared using social media.
Likes (thinking like a requester):
- Capabilities: Requesters can find capable people easily (they are not restricted by geography).
- Freelancing: Requesters don't want to hire people for their companies because is a small work.
- The experience of creating a survey: I think it can be improved because, if it's the first request for an employer or he/she just didn't understand the flow, it can frustrate him/her.
- Unknown time: My task wasn't resolved. For a requester, don't knowing the time it would take to get the job done it's stressful (maybe nobody'll take the task or the employes needs the work to be done ASAP).
This paper shows the problems a worker and a requester have when they work in a crowdworking platform and shows a way to have better experiences.
I liked so much the "boomerang", it looks like the quote 'AN EYE FOR AN EYE, AND A TOOTH FOR A TOOTH' because your decisions will affect the next works. The "prototype" was good too because its use demonstrates the kind of work the requester would be receiving and shows how an incomplete or ambiguous requirement can produce wrong results, not just the "incompetence" of a worker.
- When a worker is high-rated in an specific type of work, it can be good to prioritize him/her in a list of workers (or send him/her a message) when a similar requirement is posted. Probably these persons would have better chance to succeed and the requester would have better chances to receive a great quality HIT. I'm not suggesting to limit the access of workers to specific tasks but I think every person has its own strengths.
- It could be good to design a set of "questions/recommendations" for the recruiters analizing "well written requirements" and finding common characteristics.
- A "pro-requirements team" can be created: a recruiter can ask for the help of these teams for learning how to write requirements.
- Maybe it could be cool if, when a requester is not ranking the workers or vice versa, they receive a friendly reminder but, if they insists not doing it, they can receive some restrictions like not being able to request/take more tasks.
This paper presents a platform for low-end cellphones with iternet connection that allows people to work in OCR tasks.
- The social component: giving more people the opportunity for work in a crowdworking platform gives could make a difference in their lives: more work, more money, more food, a better life.
- The geographical component: usually a crowdworking platform worker need to be in a specific location for doing his/her work... MobileWorks allow workers to work wherever they are (if they have a normal work they can earn more money).
- Maybe the payment can be troublesome because some countries doesn't have paypal and the poorest maybe can't create a bank account.
- It would be awesome than the platform has a social media link (or something like that) for sharing the knowledge... More people could join and work then the social benefits could be greater.
This paper shows Flash Teams, teams of experts conformed by experts from a crowdworking platform (oDesk) and Foundry, a framework implementing Flash Teams.
- The possibility of having a complete team doing crowdworking because, usually, people doesn't know EVERYTHING but in a crowdworking platform, a single person takes the task or the whole project needs to be divided in not connected tasks. This is more time consuming and can be troublesome.
- More complex works can be done because, having a complete team of "experts" in specific fields makes possible to work in more complex tasks.
- Cultural differences can be hard to manage. I think maybe the algorithm can take into accound these things and prioritize more homogeneous teams when the members of a previous team had problems. The idiea is not banning heterogeneous teams, just evaluating previous experiences and, when a team experienced problems related with cultural issues, these members can be more comfortable in homogeneous teams.
- Experts recommendations can be taken into account for creating the teams, maybe their experience can be written and analized because psychological factors can be singnificatively important.
- Boomerang would be cool too: if the members of a team worked so well and they understood each other deeply, why not joining them again?