Candidate projects for homework

Project code template

Project 1: How nouns/verbs are represented in the brain? differences?

This project aims to explore how nouns and verbs are represented differently within the brain. We will utilize a neural encoding approach, employing word embeddings for nouns and verbs as inputs. The goal is to determine their correlation with functional magnetic resonance imaging (fMRI) data, revealing the distinct neural signatures of these grammatical categories.

Reference: Mitchell, T. M., Shinkareva, S. V., Carlson, A., Chang, K. M., Malave, V. L., Mason, R. A., & Just, M. A. (2008). Predicting human brain activity associated with the meanings of nouns. science, 320(5880), 1191-1195.

Project 2: How semantic relations are represented in the brain?

The focus here is on understanding how the brain encodes semantic relations, such as whole-part relationship (which could be calculated by contrasting word vectors of Finger and Hand. By calculating these relationship vectors, we will apply neural encoding methods to analyze how these semantic relations are represented neurologically, enhancing our comprehension of cognitive semantic processing.

Reference: Zhang, Y., Han, K., Worth, R., & Liu, Z. (2020). Connecting concepts in the brain by mapping cortical representations of semantic relations. Nature communications, 11(1), 1877.

Project 3: How our brain combines meaning together? i.e., which part of the brain is responsible for parsing? Which kind of parsers?

This project investigates the neural mechanisms involved in syntactic parsing—how our brain integrates meanings and identifies sentence structure. By comparing different parsing strategies (top-down, bottom-up, left-corner) and measuring syntactic complexity through node count methods, we will use neural encoding to identify the brain regions and processes responsible for parsing, providing insight into the cognitive underpinnings of language comprehension.

Reference: Zhang, X., Wang, S., Lin, N., & Zong, C. (2022, December). Is the Brain Mechanism for Hierarchical Structure Building Universal Across Languages? An fMRI Study of Chinese and English. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing (pp. 7852-7861).

Project 4: Decoding words from fMRI or EEG or MEG data (binary classification/multiple classification)

Our objective is to decode specific categories of words (e.g., animal or tool) from brain imaging data (fMRI, EEG, MEG) using classification techniques. This involves distinguishing between different categories based on neural signals, pushing the boundaries of our understanding of how language is represented in the brain.

Reference: Dirani, J., & Pylkkänen, L. (2023). The time course of cross-modal representations of conceptual categories. NeuroImage, 277, 120254.

Project 5: Decoding coherent text from fMRI data collected while listening/reading to stories

In this project, we aim to decode coherent sentences from fMRI data obtained while subjects listen to or read stories. By analyzing sequences of fMRI images for patterns associated with language processing, we intend to reconstruct the sentences, offering groundbreaking insights into real-time language comprehension and narrative processing in the brain.

Reference: Tang, J., LeBel, A., Jain, S., & Huth, A. G. (2023). Semantic reconstruction of continuous language from non-invasive brain recordings. Nature Neuroscience, 26(5), 858-866.