Turkish and Finnish annotators needed for word derivation and inflection tasks

[Mete Ismayilzada]

Dear members of SIGTURK,

We (Duygu, Jonne, Defne, and me) are working on a project to investigate the morphological generalization abilities of large language models in agglutinative languages such as Turkish and Finnish. We have designed specific word generation and recognition tasks for this purpose and would like to see how fellow humans perform on them. Please, see below for some examples of our tasks in Turkish (the Finnish examples are similar).

So, we are looking for speakers of Turkish and Finnish (preferably native) to participate in our human study. In return, we offer an authorship on our paper which we plan to submit to NAACL in two weeks. If you are interested in contributing or know someone who might, please contact me (misma...@gmail.com) soon.

Word generation task example (in Turkish)

Instruction: Given a word root and a list of affixes, construct a grammatically correct word from this word root using the given affixes.

Example: 

Word root: gerek

Affixes: se, ir

Answer: gerekirse

Word recognition task example (in Turkish)

Instruction: Given a word root, a list of affixes, and a word derived from this root using the given affixes, determine if the derived word is grammatically correct.

Example: 

Word root: okul

Affixes: unuz, a

Derivation: okulaunuz

Answer: No

Thank you very much for your time!

Best,

Mete