cjber.github.io - Cillian Berragan

Papers

Mapping Great Britain’s Semantic Footprints through a Large Language Model Analysis of Reddit Comments

open-access

github

preprint

The prevalence of social media data for use in geographic research has generated a renewed interest in the concept of ‘place’ (Wagner, Zipf, and Westerholt 2020; Purves…

Mapping Cognitive Place Associations within the United Kingdom through Online Discussion on Reddit

open-access

figshare

github

peer-reviewed

journal

This published research article builds a custom transformer-based geoparsing pipeline using the Hugging Face transformers Python library, to extract all place names from 8.3 million Reddit comments. DVC pipelines are used to ensure full reproducibility of the model workflow.

Overture POI data for the United Kingdom: a comprehensive, queryable open data product

open-access

preprint

figshare

github

Point of Interest (POI) data is an invaluable source of information, acting as a key input to much of the research that has, and continues to be generated in urban analytics…

Evaluating the Similarity of Location-based Corpora Identified in Reddit Comments

open-access

peer-reviewed

conference

Social interaction is typically studied in the context of mobility, using data sources like Census or transport records, where physical movement is restricted by distance…

Transformer based named entity recognition for place name extraction from unstructured text

open-access

figshare

github

peer-reviewed

journal

This published research article builds a custom transformer-based named entity recognition model (NER), evaluating the performance of this model against existing solutions for geoparsing geographic place names from text. This project uses the AllenNLP python library, containerised using Docker for reproducibility.

Geoparsing comments from Reddit to extract mental place connectivity within the United Kingdom

open-access

peer-reviewed

conference

Connectivity between places may be explored through the physical movement of individuals, using population movement data like transport records (Yang, Li, and Li 2019; Allard…

Comparing rule-based methods and pre-trained language models to classify flood related Tweets

open-access

peer-reviewed

conference

Twitter presents large continuous feed of information regarding emergency events, contributed through individual users, as these events occur. Many emergency events have…