Text related annotation
Verify 3 question–answer pairs for each image. For each pair: (1) mark whether the Q/A is correct and relevant to the image, and (2) mark whether the question cannot be answered without seeing the image.
NativQA Framework Answer Editing - Egypt (Cairo)
NativQA Framework Answer Editing - Egypt (Eg)
NativQA Framework Answer Editing - Mauritania
NativQA Framework Answer Editing - Palestanian
NativQA Framework Answer Editing - Saudi Arabia
Read and record English questions for the SemEval 2027 multimodal dataset.
Read and record Modern Standard Arabic questions for the SemEval 2027 multimodal dataset.