Course at UNITN Italy

Neurolinguistics and conceptual modeling in neuroimaging (2017-2018 1e semestre PSICOLOGIA-Psicologia delle risorse umane e organizzazione) at University of Trento, Italy

(Tutte le lezioni all'Università di Trento sono finite. Grazie mille.)

Time Table

1 2017/11/20 MON 17:00-19:00

2 2017/11/21 TUE 17:00-19:00

3 2017/11/22 WED 15:00-17:00

4 2017/11/27 MON 17:00-19:00

5 2017/11/28 TUES 17:00-19:00

6 2017/11/29 WED 15:00-17:00

7 2017/12/1 FRI 15:00-17:00

8 2017/12/4 MON 17:00-19:00

9 2017/12/5 TUE 17:00-19:00

10 2017/12/6 WED 15:00-17:00

11 2017/12/11 MON 17:00-19:00

12 2017/12/12 TUE 17:00-19:00

13 2017/12/13 WED 15:00-17:00

14 2017/12/15 FRI 15:00-17:00

Syllabus of the course: Neurolinguistics and conceptual modeling in neuroimaging

Lecturer: Hiroyuki Akama, Visiting Professor at the University of Trento, Department of Psychology and Cognitive Science.

 Contents

              This course focuses on neurolinguistics and covers the fundamentals of i) embodied cognition and neuroimaging, ii) computational neurolinguistics by machine learning, and iii) semantic network and graph theoretical analysis in fMRI (default mode network). Embodiment theory, which plays an important role in neurocognitive linguistics, posits the mechanism of semantic processing in a general context of body and mind, action and perception, movement and imagery. In this regard, recordings of neuroimaging data by using functional magnetic resonance imaging (fMRI) should allow us to understand how the mind and brain represent and process concepts, as they are instantiated in particular human languages. This comprehensive approach has the potential for substantializing symbolic and/or linguistic encodings when based on computational analyses to probe human neural data. Computational neurolinguistics can be considered crucial in that respect as it embraces various human cognitive functions in an estimable model by integrating computational linguistics and cognitive neuroscience, especially taking advantage of machine learning methodologies. Furthermore, we will emphasize in this context the applicability of emerging graph theoretical techniques to the semantic network as related to one of the most typical brain network called "default mode" which has been revealed by the resting-state functional connectivity MRI. 

              Altogether, this course will target the comprehensive modeling of language processing based on fMRI data, and will emphasize the embodied encoding of meaning in the brain and computational language processing with the view of coordinating these seemingly heterogeneous research fields into a future perspective of language science.

Objectives

              At the end of this course, students will be able to:

i) Explain the concept of neurolinguistics and conceptual modeling in neuroimaging.

ii) Prepare an overview of previously conducted research, identify issues and points of controversy, and propose a solution.

iii) Design an experiment or draw up a future research plan pertaining to cognitive semantics and other relevant domains.

              Students will have the chance to tackle theoretical and/or practical problems in neurocognitive science by applying knowledge acquired through this course.

Prerequisites

               Registration for the course requires no particular prerequisites, but some familiarity with brain and language. A beginner level of programming skill might be effective for taking advantage of the contents provided during this course. Students are requested to bring their own laptop computers or smart phones in order to perform information retrieval upon brain atlases available on the web.

Teaching Methods

              Before coming to class, students should check what topics will be covered that day. They should read the course materials (pdf version of PowerPoint slides that the lecturer created and uploaded to his official web site) and at least the abstracts of the journal papers enumerated in the syllabus and the course materials. Required learning should be completed outside of the classroom for preparation and review purposes.

             At the beginning of each class, the lecture outlines are presented to introduce the essence of neurolinguistics and fMRI by using the course materials. Students are expected to discuss key lines of research by reading some journal papers representing established or cutting-edge studies in this field. A framework of active learning will be exploited for collaboration between the lecturer and the students for brainstorming and mutual review of ideas. It will be also enhanced by individual computer training for intelligent retrieval of multi-angle information from the internet. In particular, students will get used to the manipulation of the brain datasets shared through the sites of fMRI meta-analysis (such as Neurosynth, Neurovault, Brede database, etc.) to correlate the location of the brain regions which are informative of the features treated in the fMRI papers.

Verification of learning

             Students will be assessed on their understanding of the aforementioned fundamentals. The evaluation reflects the quality of students’ performance in the class. Their course scores are based on participation in journal paper readings, the quizzes after each session, and the final exam or report.

 More Information

              About the lecturer, refer to the following URLs.

http://educ.titech.ac.jp/bio/eng/news/2017_03/053504.html

https://www.researchgate.net/profile/Hiroyuki_Akama

http://orcid.org/0000-0003-1777-497X

Texts

All materials used in class can be found on the official web site of the lecturer.

The journal papers treated in class are the following but not limited to

i) embodied cognition and neuroimaging

1      Friedemann Pulvermüller, Brainmechanisms linking language and action, Nature, 2005. doi:10.1038/nrn1706

2      Roel M. Willems and DanielCasasanto, Flexibility in embodied language understanding, frontiers inpsychology, 2011. doi: 10.3389/fpsyg.2011.00116

3      Scott L. Fairhall and AlfonsoCaramazza, Brain Regions That Represent Amodal Conceptual Knowledge, TheJournal of Neuroscience, 2013. 33(25):10552–10558

4      Giovanni Pezzulo et al.,Computational Grounded Cognition: a new alliance between grounded cognition andcomputational modeling, frontiers in Psychology, 2013.  https://doi.org/10.3389/fpsyg.2012.00612

5      Lawrence W. Barsalou, What doessemantic tiling of the cortex tell us about semantics?, Neuropsychologia,(Article in press)

ii) computational neurolinguistics by machine learning

1      Tom M. Mitchell, et al.,Predicting Human Brain Activity Associated with the Meanings of Nouns, Science,2008. DOI: 10.1126/science.1152876

2      John A. Bullinaria et al.,Limiting Factors for Mapping Corpus-Based Semantic Representations to BrainActivity, PLoS ONE, 2013. doi:10.1371/journal.pone.0057191

3      Andrew James Anderson et al.,Predicting Neural Activity Patterns Associated with Sentences Using aNeurobiologically Motivated Model of Semantic Representation, Cerebral Cortex,2016; 1–17 doi: 10.1093/cercor/bhw240

4      Hiroyuki Akama et al., UsingGraph Components Derived from an Associative Concept Dictionary to Predict fMRINeural Activation Patterns that Represent the Meaning of Nouns, PLoS ONE, 2015.DOI: 10.1371/journal.pone.0125725

iii) semantic network and graph theoretical analysis in fMRI

1.     Olaf Sporns et al.,Organization, development and function of complex brain networks, Trends inCognitive Science, 2004. Vol.8 No.9 pp.871–882.

2.     W. Dale Stevens and R. NathanSpreng, Resting-state functional connectivity MRI reveals active processescentral to cognition, Wiley Interdisciplinary Reviews: Cognitive Science, DOI: 10.1002/wcs.1275

3.     Jeffrey R. Binder and Rutvik H.Desai, The neurobiology of semantic memory, Trends in Cognitive Science, 2011.https://doi.org/10.1016/j.tics.2011.10.001

4.     Gina F. Humphreys et al.,Establishing task- and modality-dependent dissociations between the semanticand default mode networks, PNAS, 2015. doi: 10.1073/pnas.1422760112.