Hello Everyone, We are currently publishing a set of some exciting HITs on AMT. They require your judgement of text given some data. You are asked to evaluate the text across multiple criteria given the data. Your judgement should be faithful and reasonable. Of course, because of the nature of the task, your judgements might be subjective (in fact, they should be), but it's ok if so far they are more or less reasonable. The HITs can be found by looking for "Rate Correspondence Between Data and Text" or "Claire Gardent CNRS/LORIA" in the search field. There are two types of HITs: 1. Qualification HIT: first, you are required to complete the qualification HIT. This HIT is used to ensure that you complete your work faithfully. We filter malicious workers this way. The HITs are named Rate Correspondence Between Data and Text [qualification HIT]. 2. Original HIT: if you pass our qualification HIT, you will be able to work on the actual HITs. You can complete around 70 of these HITs. However, this week we are planning to run more of these HITs, where workers are allowed to complete even more HITs (100 and more). The HITs are named Rate Correspondence Between Data and Text [original HIT]. We approve HITs in max one day and always quick with responding. We would be grateful if you could find some time to work on our tasks. If you have questions, you are welcome to ask me =) Best, Nikolai.
