MODDP: A Multi-modal Open-domain Chinese Dataset for Dialogue Discourse Parsing.
Chen GongDexin KongSuxian ZhaoXingyu LiGuohong FuPublished in: ACL (Findings) (2024)
Keyphrases
- multi modal
- open domain
- natural language
- information extraction
- question answering
- discourse structure
- unknown words
- speech acts
- multi modality
- natural language processing
- question answering systems
- high dimensional
- image annotation
- audio visual
- uni modal
- video search
- cross modal
- machine learning
- co occurrence
- word segmentation
- information retrieval