With no explanation, label text_A→text_B with either "strengthener" or "weakener".
text_A: This man is smiling very big at the camera. The drunk man looks at the video camera.
text_B: The man is holding a beer bottle in his hand
strengthener.