ChineseEEG: A Chinese Linguistic Corpora EEG Dataset for Semantic Alignment and Neural Decoding

Area of Science:

• Neuroscience
• Cognitive Science
• Linguistics

Background:

• Understanding brain's semantic encoding is crucial for brain-computer interfaces (BCI).
• Existing Electroencephalography (EEG) datasets lack Chinese linguistic stimuli.
• There is a need for resources to study brain mechanisms of Chinese language processing.

Purpose of the Study:

• To introduce the ChineseEEG dataset, a high-density EEG dataset with simultaneous eye-tracking.
• To provide a resource for studying semantic decoding using natural Chinese text.
• To facilitate research in neuroscience, natural language processing (NLP), and linguistics.

Main Methods:

• Recorded high-density EEG and eye-tracking from 10 participants reading Chinese novels for ~13 hours.
• Collected long-duration EEG data during naturalistic reading.
• Extracted semantic embeddings using a pre-trained NLP model.

Main Results:

• The ChineseEEG dataset includes pre-processed EEG sensor-level data.
• Semantic embeddings of the reading materials are provided.
• The dataset offers a benchmark for Chinese semantic decoding.

Conclusions:

• ChineseEEG is a valuable pilot dataset for Chinese natural language processing and cognitive research.
• It supports the development of BCIs and exploration of large language model alignment with human cognition.
• Facilitates research into the neural basis of Chinese language comprehension.