Mordecai 3: A Neural Geoparser and Event Geocoder

TL;DR

Mordecai 3 introduces an end-to-end neural geoparser and event geocoder that improves toponym resolution and event linking in text, leveraging neural ranking and question-answering models, and is available as an open source tool.

Contribution

It presents a novel neural ranking model for toponym resolution and integrates event geocoding using question-answering, enhancing existing geoparsing capabilities.

Findings

Outperforms existing geoparsers in toponym resolution accuracy.

Successfully links events to locations in text.

Open source implementation available for community use.

Abstract

Mordecai3 is a new end-to-end text geoparser and event geolocation system. The system performs toponym resolution using a new neural ranking model to resolve a place name extracted from a document to its entry in the Geonames gazetteer. It also performs event geocoding, the process of linking events reported in text with the place names where they are reported to occur, using an off-the-shelf question-answering model. The toponym resolution model is trained on a diverse set of existing training data, along with several thousand newly annotated examples. The paper describes the model, its training process, and performance comparisons with existing geoparsers. The system is available as an open source Python library, Mordecai 3, and replaces an earlier geoparser, Mordecai v2, one of the most widely used text geoparsers (Halterman 2017).