{"id":{"repo_id":"birmingham","oai_identifier":"oai:etheses.bham.ac.uk:12966"},"canonical_url":"https://search.ndltd.org/etd/birmingham/oai:etheses.bham.ac.uk:12966","repository":{"repo_id":"birmingham","name":"University of Birmingham","base_url":"https://etheses.bham.ac.uk/cgi/oai2"},"display":{"title":"Decoding linguistic information from EEG signals","abstract":"For many years, the fields of the cognitive neuroscience of language and natural language processing (NLP) have been relatively distinct and non-overlapping. Recent breakthrough research is starting to show that these two fields, in their common goal towards understanding and modelling language, have a lot to offer each other. As developments in machine learning continue to break into new ground, due largely in part to the successful development of novel classifiers that can be efficiently trained to model highly nonlinear dynamic systems, such as language, the open question is how well these models perform on human neural signals during language processing. Recent results are beginning to show that various types of human signals (eye-tracking, fMRI, MEG) can successfully model various linguistic aspects of what is being concurrently processed by the brain. EEG is a cheap and relatively accessible way to access neural signals and this thesis explores the extent to which decoding of EEG data, using state-of-the-art models common in NLP, to carry out this task. Critically, an important foundation needs to be in place that can fully explore the types of linguistic signal that is decodable with EEG. This thesis attempts to answer this question, setting the stage for joint modelling of text and neural signals to advance the field of NLP. This research is also of interest to cognitive neuroscientists as the data collected for this thesis will be openly accessible to all, with accompanying linguistic annotation, which can help to answer various questions about the spatiotemporal dynamics during the reading of naturalistic texts. In Chapter 1, I provide an overview of the major literature that has investigated the status of linguistic processing from neural signals, setting the research question in the correct historical context. This literature review serves as the basis for the two experimental chapters which follow and is thus subdivided into two main sections. Chapter 2 explores the various aspects of linguistic processing which are decodable from the novel EEG dataset collected for this thesis, with a strong emphasis on controlling for potential confounds as much as possible. Using a novel machine learning classifier, I show that with specialised training methods, generalisation to novel data relating to part-of-speech decoding is possible. In Chapter 3, the preprocessing steps involved in preparing the data are examined, in which I show that depending on the modelling goal, some steps are particularly useful to boost performance of linguistic decoding of EEG stimuli. Finally, in Chapter 4, a broad review of the results, their implications and limitations are considered.","abstract_html":"For many years, the fields of the cognitive neuroscience of language and natural language processing (NLP) have been relatively distinct and non-overlapping. Recent breakthrough research is starting to show that these two fields, in their common goal towards understanding and modelling language, have a lot to offer each other. As developments in machine learning continue to break into new ground, due largely in part to the successful development of novel classifiers that can be efficiently trained to model highly nonlinear dynamic systems, such as language, the open question is how well these models perform on human neural signals during language processing. Recent results are beginning to show that various types of human signals (eye-tracking, fMRI, MEG) can successfully model various linguistic aspects of what is being concurrently processed by the brain. EEG is a cheap and relatively accessible way to access neural signals and this thesis explores the extent to which decoding of EEG data, using state-of-the-art models common in NLP, to carry out this task. Critically, an important foundation needs to be in place that can fully explore the types of linguistic signal that is decodable with EEG. This thesis attempts to answer this question, setting the stage for joint modelling of text and neural signals to advance the field of NLP. This research is also of interest to cognitive neuroscientists as the data collected for this thesis will be openly accessible to all, with accompanying linguistic annotation, which can help to answer various questions about the spatiotemporal dynamics during the reading of naturalistic texts. In Chapter 1, I provide an overview of the major literature that has investigated the status of linguistic processing from neural signals, setting the research question in the correct historical context. This literature review serves as the basis for the two experimental chapters which follow and is thus subdivided into two main sections. Chapter 2 explores the various aspects of linguistic processing which are decodable from the novel EEG dataset collected for this thesis, with a strong emphasis on controlling for potential confounds as much as possible. Using a novel machine learning classifier, I show that with specialised training methods, generalisation to novel data relating to part-of-speech decoding is possible. In Chapter 3, the preprocessing steps involved in preparing the data are examined, in which I show that depending on the modelling goal, some steps are particularly useful to boost performance of linguistic decoding of EEG stimuli. Finally, in Chapter 4, a broad review of the results, their implications and limitations are considered.","abstract_has_math":false,"creators":["Murphy, Alex"],"institution":"University of Birmingham","degree_name":"d_ph","degree_level":"d_ph","degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2022,"date_issued":"2022-12","date_published":"2022-12","updated_at":"2026-11-19T09:32:07Z","subjects":["AI Indexes (General)","BF Psychology","QA75 Electronic computers. Computer science","QP Physiology"],"languages":["eng"],"rights":[],"rights_urls":[],"identifier_entries":[],"person_identifier_entries":[{"key":"person:identifier.orcid","label":"ORCID","values":["0000000161550514"],"render_values":[{"text":"0000-0001-6155-0514","href":"https://orcid.org/0000-0001-6155-0514","code":true}]}]},"links":{"outbound_url":null,"outbound_label":null,"outbound_source":null,"record_url":null,"record_label":null,"record_kind":null,"record_source":null,"download_url":"http://etheses.bham.ac.uk//id/eprint/12966/7/Murphy2022PhD.pdf","download_label":"Download PDF","download_source":"dc:identifier.uri","access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://etheses.bham.ac.uk/cgi/oai2?verb=GetRecord&metadataPrefix=uketd_dc&identifier=oai%3Aetheses.bham.ac.uk%3A12966","prefix":"uketd_dc"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.sponsor","label":"Sponsor","values":["bbsrc"]},{"key":"dc:creator","label":"Author","values":["Murphy, Alex"]},{"key":"person:identifier.orcid","label":"ORCID","values":["0000000161550514"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2022-12-05"]},{"key":"dc:date.issued","label":"Date","values":["2022-12"]},{"key":"dc:publisher.department","label":"Dc Publisher Department","values":["College of Life & Environmental Sciences","School of Psychology"]},{"key":"dc:publisher.institution","label":"Dc Publisher Institution","values":["University of Birmingham"]},{"key":"dc:relation.isreferencedby","label":"Dc Relation Isreferencedby","values":["http://etheses.bham.ac.uk//id/eprint/12966/"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"dc:type.qualificationlevel","label":"Dc Type Qualificationlevel","values":["d_ph"]},{"key":"dc:type.qualificationname","label":"Dc Type Qualificationname","values":["d_ph"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["AI Indexes (General)","BF Psychology","QA75 Electronic computers. Computer science","QP Physiology"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.dcterms_iso639_2","label":"Language (ISO 639-2)","values":["eng"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["http://etheses.bham.ac.uk//id/eprint/12966/7/Murphy2022PhD.pdf"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["For many years, the fields of the cognitive neuroscience of language and natural language processing (NLP) have been relatively distinct and non-overlapping. Recent breakthrough research is starting to show that these two fields, in their common goal towards understanding and modelling language, have a lot to offer each other. As developments in machine learning continue to break into new ground, due largely in part to the successful development of novel classifiers that can be efficiently trained to model highly nonlinear dynamic systems, such as language, the open question is how well these models perform on human neural signals during language processing. Recent results are beginning to show that various types of human signals (eye-tracking, fMRI, MEG) can successfully model various linguistic aspects of what is being concurrently processed by the brain. EEG is a cheap and relatively accessible way to access neural signals and this thesis explores the extent to which decoding of EEG data, using state-of-the-art models common in NLP, to carry out this task. Critically, an important foundation needs to be in place that can fully explore the types of linguistic signal that is decodable with EEG. This thesis attempts to answer this question, setting the stage for joint modelling of text and neural signals to advance the field of NLP. This research is also of interest to cognitive neuroscientists as the data collected for this thesis will be openly accessible to all, with accompanying linguistic annotation, which can help to answer various questions about the spatiotemporal dynamics during the reading of naturalistic texts. In Chapter 1, I provide an overview of the major literature that has investigated the status of linguistic processing from neural signals, setting the research question in the correct historical context. This literature review serves as the basis for the two experimental chapters which follow and is thus subdivided into two main sections. Chapter 2 explores the various aspects of linguistic processing which are decodable from the novel EEG dataset collected for this thesis, with a strong emphasis on controlling for potential confounds as much as possible. Using a novel machine learning classifier, I show that with specialised training methods, generalisation to novel data relating to part-of-speech decoding is possible. In Chapter 3, the preprocessing steps involved in preparing the data are examined, in which I show that depending on the modelling goal, some steps are particularly useful to boost performance of linguistic decoding of EEG stimuli. Finally, in Chapter 4, a broad review of the results, their implications and limitations are considered."]},{"key":"dc:format","label":"Dc Format","values":["text"]},{"key":"dcterms:accessrights","label":"Dcterms Accessrights","values":["public"]},{"key":"dc:title","label":"Title","values":["Decoding linguistic information from EEG signals"]}]}],"canonical_facts":{"dc:contributor.sponsor":["bbsrc"],"dc:creator":["Murphy, Alex"],"dc:date":["2022-12-05"],"dc:date.issued":["2022-12"],"dc:description.abstract":["For many years, the fields of the cognitive neuroscience of language and natural language processing (NLP) have been relatively distinct and non-overlapping. Recent breakthrough research is starting to show that these two fields, in their common goal towards understanding and modelling language, have a lot to offer each other. As developments in machine learning continue to break into new ground, due largely in part to the successful development of novel classifiers that can be efficiently trained to model highly nonlinear dynamic systems, such as language, the open question is how well these models perform on human neural signals during language processing. Recent results are beginning to show that various types of human signals (eye-tracking, fMRI, MEG) can successfully model various linguistic aspects of what is being concurrently processed by the brain. EEG is a cheap and relatively accessible way to access neural signals and this thesis explores the extent to which decoding of EEG data, using state-of-the-art models common in NLP, to carry out this task. Critically, an important foundation needs to be in place that can fully explore the types of linguistic signal that is decodable with EEG. This thesis attempts to answer this question, setting the stage for joint modelling of text and neural signals to advance the field of NLP. This research is also of interest to cognitive neuroscientists as the data collected for this thesis will be openly accessible to all, with accompanying linguistic annotation, which can help to answer various questions about the spatiotemporal dynamics during the reading of naturalistic texts. In Chapter 1, I provide an overview of the major literature that has investigated the status of linguistic processing from neural signals, setting the research question in the correct historical context. This literature review serves as the basis for the two experimental chapters which follow and is thus subdivided into two main sections. Chapter 2 explores the various aspects of linguistic processing which are decodable from the novel EEG dataset collected for this thesis, with a strong emphasis on controlling for potential confounds as much as possible. Using a novel machine learning classifier, I show that with specialised training methods, generalisation to novel data relating to part-of-speech decoding is possible. In Chapter 3, the preprocessing steps involved in preparing the data are examined, in which I show that depending on the modelling goal, some steps are particularly useful to boost performance of linguistic decoding of EEG stimuli. Finally, in Chapter 4, a broad review of the results, their implications and limitations are considered."],"dc:format":["text"],"dc:identifier.uri":["http://etheses.bham.ac.uk//id/eprint/12966/7/Murphy2022PhD.pdf"],"dc:language.dcterms_iso639_2":["eng"],"dc:publisher.department":["College of Life & Environmental Sciences","School of Psychology"],"dc:publisher.institution":["University of Birmingham"],"dc:relation.isreferencedby":["http://etheses.bham.ac.uk//id/eprint/12966/"],"dc:subject":["AI Indexes (General)","BF Psychology","QA75 Electronic computers. Computer science","QP Physiology"],"dcterms:accessrights":["public"],"dc:title":["Decoding linguistic information from EEG signals"],"dc:type":["Thesis"],"dc:type.qualificationlevel":["d_ph"],"dc:type.qualificationname":["d_ph"],"person:identifier.orcid":["0000000161550514"]},"updated_at":"2026-11-19T09:32:07Z"}