geoparsepy project

geoparsepy is a Python geoparsing library that will extract and disambiguate locations from text. It uses a local OpenStreetMap database which allows very high and unlimited geoparsing throughput, unlike approaches that use a third-party geocoding service (e.g. Google Geocoding API).

geoparsepy PyPI

Geoparsing is based on named entity matching against OpenStreetMap (OSM) locations. All locations with names that match tokens will be selected from a target text sentence. This will result in a set of OSM locations, all with a common name or name variant, for each token in the text. Geoparsing included the following features:

• token expansion using location name variants (i.e. OSM multi-lingual names, short names and acronyms)
• token expansion using location type variants (e.g. street, st.)
• token filtering of single token location names against WordNet (non-nouns), language specific stoplists and peoples first names (nltk.corpus.names.words()) to reduce false positive matches
• prefix checking when matching in case a first name prefixes a location token(s) to avoid matching peoples full names as locations (e.g. Victoria Derbyshire != Derbyshire)

Location disambiguation is the process of choosing which of a set of possible OSM locations, all with the same name, is the best match. Location disambiguation is based on an evidential approach, with evidential features detailed below in order of importance:

• token subsumption, rejecting smaller phrases over larger ones (e.g. 'New York' will prefer [New York, USA] to [York, UK])
• nearby parent region, preferring locations with a parent region also appearing within a semantic distance (e.g. 'New York in USA' will prefer [New York, USA] to [New York, BO, Sierra Leone])
• nearby locations, preferring locations with closeby or overlapping locations within a semantic distance (e.g. 'London St and Commercial Road' will select from road name choices with the same name based on spatial proximity)
• nearby geotag, preferring locations that are closeby or overlapping a geotag
• general before specific, rejecting locations with a higher admin level (or no admin level at all) compared to locations with a lower admin level (e.g. 'New York' will prefer [New York, USA] to [New York, BO, Sierra Leone]

Currently the following languages are supported:

• English, French, German, Italian, Portuguese, Russian, Ukrainian
• All other languages will work but there will be no language specific token expansion available

geoparsepy works with Python 3.7 and has been tested on Windows 10 and Ubuntu 18.04 LTS.

This geoparsing algorithm uses a large memory footprint (e.g. 12 Gbytes RAM for global cities), RAM size proportional to the number of cached locations, to maximize matching speed. It can be naively parallelized, with multiple geoparse processes loaded with different sets of locations and the geoparse results aggregated in a last process where location disambiguation is applied. This approach has been validated across an APACHE Storm cluster.

The software is copyright 2020 University of Southampton, UK. It was created over a multi-year period under EU FP7 projects TRIDEC (258723), REVEAL (610928), InnovateUK project LPLP (104875) and ESRC project FloraGuard (ES/R003254/1). This software can only be used for research, education or evaluation purposes. A free commercial license is available on request to {sem03}@soton.ac.uk. The University of Southampton is open to discussions regarding collaboration in future research projects relating to this work.

Feature suggestions and/or bug reports can be sent to {sem03}@soton.ac.uk. We do not however offer any software support beyond the examples and API documentation already provided.

Scientific publications

Middleton, S.E. Middleton, L. Modafferi, S. Real-time Crisis Mapping of Natural Disasters using Social Media, Intelligent Systems, IEEE , vol.29, no.2, pp.9,17, Mar.-Apr. 2014

Middleton, S.E. Krivcovs, V. Geoparsing and Geosemantics for Social Media: Spatio-Temporal Grounding of Content Propagating Rumours to support Trust and Veracity Analysis during Breaking News, ACM Transactions on Information Systems (TOIS), 34, 3, Article 16 (April 2016), 26 pages. DOI=10.1145/2842604

Middleton, S.E. Kordopatis-Zilos, G. Papadopoulos, S. Kompatsiaris, Y. Location Extraction from Social Media: Geoparsing, Location Disambiguation, and Geotagging, ACM Transactions on Information Systems (TOIS) 36, 4, Article 40 (June 2018), 27 pages. DOI: https://doi.org/10.1145/3202662. Presented at SIGIR 2019

A benchmark geoparse dataset is also available for free from the University of Southampton on request via email to {sem03}@soton.ac.uk.

geoparsepy documentation resources

geoparsepy API

geoparsepy example code on github

Python libs needed (earlier versions may be suitable but are untested)

Python libs: psycopg2 >= 2.8, nltk >= 3.4, numpy >= 1.18, shapely >= 1.6, setuptools >= 46, soton-corenlppy>=1.0

Database: PostgreSQL >= 11.3, PostGIS >= 2.5

For LINUX deployments the following is needed:

sudo apt-get install libgeos-dev libgeos-3.4.2 libpq-dev

You will need to download NLTK corpra before running geoparsepy:

python
import nltk
nltk.download()
==> install all or at least stopwords, names and wordnet

Installation

python3 -m pip install geoparsepy

Databases needed for geoparsing

Download pre-processed UTF-8 encoded SQL table dumps from OSM image dated dec 2019. SQL dump is a 1.2 GB tar/zip file created using pg_dump and zipped using 7Zip tool.

download zip file from Google drive https://drive.google.com/file/d/1xyCjQox6gCoN8e0upHHyeMLV-uLirthS/view?usp=sharing
unzip geoparsepy_preprocessed_tables.tar.zip
tar -xvf geoparsepy_preprocessed_tables.tar

Connect to PostgreSQL and create the database with the required PostGIS and hstore extensions

psql -U postgres
CREATE DATABASE openstreetmap;
CREATE EXTENSION IF NOT EXISTS postgis;
CREATE EXTENSION IF NOT EXISTS fuzzystrmatch;
CREATE EXTENSION IF NOT EXISTS postgis_tiger_geocoder;
CREATE EXTENSION IF NOT EXISTS hstore;

Import the precomputed database tables for global cities and places

# Linux
psql -U postgres -d openstreetmap -f global_cities.sql
psql -U postgres -d openstreetmap -f uk_places.sql
psql -U postgres -d openstreetmap -f north_america_places.sql
psql -U postgres -d openstreetmap -f europe_places.sql

# Windows 10 using powershell
& 'C:\Program Files\PostgreSQL\11\bin\psql.exe' -U postgres -d openstreetmap -f global_cities.sql
& 'C:\Program Files\PostgreSQL\11\bin\psql.exe' -U postgres -d openstreetmap -f uk_places.sql
& 'C:\Program Files\PostgreSQL\11\bin\psql.exe' -U postgres -d openstreetmap -f north_america_places.sql
& 'C:\Program Files\PostgreSQL\11\bin\psql.exe' -U postgres -d openstreetmap -f europe_places.sql

# Linux if username not sem (used for sql dump) not added as a user and some other name is needed (e.g. sem03)
find * -name \*.sql -exec sed -i "s/TO sem;/TO sem03;/g" {} \;
psql -d openstreetmap -f uk_places.sql
psql -d openstreetmap -f global_cities.sql
psql -d openstreetmap -f north_america_places.sql
psql -d openstreetmap -f europe_places.sql

Example code geoparse (start here)

Geoparse some text using the default focus areas in the Postgres database. Fully documented example PY file can be found at geoparsepy.example_geoparse.py note: loading 300,000+ global locations into memory at startup is slow (10 minutes) but subsequently the geoparsing of text is very fast (real-time speeds)

import os, sys, logging, traceback, codecs, datetime, copy, time, ast, math, re, random, shutil, json
import soton_corenlppy, geoparsepy

LOG_FORMAT = ('%(message)s')
logger = logging.getLogger( __name__ )
logging.basicConfig( level=logging.INFO, format=LOG_FORMAT )
logger.info('logging started')

dictGeospatialConfig = geoparsepy.geo_parse_lib.get_geoparse_config( 
	lang_codes = ['en'],
	logger = logger,
	whitespace = u'"\u201a\u201b\u201c\u201d()',
	sent_token_seps = ['\n','\r\n', '\f', u'\u2026'],
	punctuation = """,;\/:+-#~&*=!?""",
	)

databaseHandle = soton_corenlppy.PostgresqlHandler.PostgresqlHandler( 'postgres', 'postgres', 'localhost', 5432, 'openstreetmap', 600 )

dictLocationIDs = {}
listFocusArea=[ 'global_cities', 'europe_places', 'north_america_places', 'uk_places' ]
for strFocusArea in listFocusArea :
	dictLocationIDs[strFocusArea + '_admin'] = [-1,-1]
	dictLocationIDs[strFocusArea + '_poly'] = [-1,-1]
	dictLocationIDs[strFocusArea + '_line'] = [-1,-1]
	dictLocationIDs[strFocusArea + '_point'] = [-1,-1]

cached_locations = geoparsepy.geo_preprocess_lib.cache_preprocessed_locations( databaseHandle, dictLocationIDs, 'public', dictGeospatialConfig )
logger.info( 'number of cached locations = ' + str(len(cached_locations)) )

databaseHandle.close()

indexed_locations = geoparsepy.geo_parse_lib.calc_inverted_index( cached_locations, dictGeospatialConfig )
logger.info( 'number of indexed phrases = ' + str(len(indexed_locations.keys())) )

indexed_geoms = geoparsepy.geo_parse_lib.calc_geom_index( cached_locations )
logger.info( 'number of indexed geoms = ' + str(len(indexed_geoms.keys())) )

osmid_lookup = geoparsepy.geo_parse_lib.calc_osmid_lookup( cached_locations )

dictGeomResultsCache = {}

listText = [
	u'hello New York, USA its Bill from Bassett calling',
	u'live on the BBC Victoria Derbyshire is visiting Derbyshire for an exclusive UK interview',
	]

listTokenSets = []
listGeotags = []
for nIndex in range(len(listText)) :
	strUTF8Text = listText[ nIndex ]
	listToken = soton_corenlppy.common_parse_lib.unigram_tokenize_text( text = strUTF8Text, dict_common_config = dictGeospatialConfig )
	listTokenSets.append( listToken )
	listGeotags.append( None )

listMatchSet = geoparsepy.geo_parse_lib.geoparse_token_set( listTokenSets, indexed_locations, dictGeospatialConfig )

strGeom = 'POINT(-1.4052268 50.9369033)'
listGeotags[0] = strGeom

listMatchGeotag = geoparsepy.geo_parse_lib.reverse_geocode_geom( [strGeom], indexed_geoms, dictGeospatialConfig )
if len( listMatchGeotag[0] ) > 0  :
	for tupleOSMIDs in listMatchGeotag[0] :
		setIndexLoc = osmid_lookup[ tupleOSMIDs ]
		for nIndexLoc in setIndexLoc :
			strName = cached_locations[nIndexLoc][1]
			logger.info( 'Reverse geocoded geotag location [index ' + str(nIndexLoc) + ' osmid ' + repr(tupleOSMIDs) + '] = ' + strName )

for nIndex in range(len(listMatchSet)) :
	logger.info( 'Text = ' + listText[nIndex] )
	listMatch = listMatchSet[ nIndex ]
	strGeom = listGeotags[ nIndex ]
	setOSMID = set([])
	for tupleMatch in listMatch :
		nTokenStart = tupleMatch[0]
		nTokenEnd = tupleMatch[1]
		tuplePhrase = tupleMatch[3]
		for tupleOSMIDs in tupleMatch[2] :
			setIndexLoc = osmid_lookup[ tupleOSMIDs ]
			for nIndexLoc in setIndexLoc :
				logger.info( 'Location [index ' + str(nIndexLoc) + ' osmid ' + repr(tupleOSMIDs) + ' @ ' + str(nTokenStart) + ' : ' + str(nTokenEnd) + '] = ' + ' '.join(tuplePhrase) )
				break
	listLocMatches = geoparsepy.geo_parse_lib.create_matched_location_list( listMatch, cached_locations, osmid_lookup )
	geoparsepy.geo_parse_lib.filter_matches_by_confidence( listLocMatches, dictGeospatialConfig, geom_context = strGeom, geom_cache = dictGeomResultsCache )
	geoparsepy.geo_parse_lib.filter_matches_by_geom_area( listLocMatches, dictGeosp