=Paper=
{{Paper
|id=Vol-3385/paper5
|storemode=property
|title=The Koja Web Mapping Application for Context-sensitive Natural Language Spatial Querying
|pdfUrl=https://ceur-ws.org/Vol-3385/paper5.pdf
|volume=Vol-3385
|authors=Niloofar Aflaki,Kristin Stock,Christopher B. Jones,Hans Guesgen,Yukio Fukuzawa,Jeremy Morley
|dblpUrl=https://dblp.org/rec/conf/ecir/AflakiSJGFM23
}}
==The Koja Web Mapping Application for Context-sensitive Natural Language Spatial Querying==
https://ceur-ws.org/Vol-3385/paper5.pdf
The Koja Web Mapping Application for
Context-sensitive Natural Language Spatial Querying
Niloofar Aflaki1,∗ , Kristin Stock1 , Christopher B. Jones2 , Hans Guesgen1 ,
Yukio Fukuzawa1 and Jeremy Morley3
1
Massey Geoinformatics Collaboratory, Massey University, Auckland, New Zealand
2
Cardiff University, United Kingdom
3
Ordnance Survey, United Kingdom
Abstract
The locations of objects are often described in natural language relative to some other object using
vague and context-sensitive spatial relation terms (e.g. theatre near Trafalgar Square). Koja is a web
map application that predicts the distance between a location and reference object based on the spatial
relation term specified by the user and language and contextual features. That distance is used to retrieve
objects of the specified type within a range of the distance. They are displayed through a map interface
to make the process more intuitive and user-friendly.
Keywords
Web map, location description, linguistic features, spatial relation term, spatial preposition
1. Introduction
In natural language, people often describe location using locative expressions [1] that include
three key elements: the place or object the location of which is being described (known as the
locatum or trajector), a place which is the reference object (known as the relatum or landmark)
and a relationship between these two (spatial relation term, often a preposition), e.g., theatre
[locatum] near [spatial relation term] Trafalgar Square [relatum]. Koja1 is an interactive web
map which retrieves a set of candidate objects that correspond to the described locatum type,
using machine learning to predict the distance between the locatum and relatum based on a
range of linguistic and contextual features from the three elements in a given expression.
Web maps have been created to answer a wide range of research questions. Environmental
applications such as ClimateCharts, which provides an interactive web map showing temper-
ature and precipitation of most places on earth [2], are common, and enable researchers to
explore various physical interactions. Web mapping applications that incorporate individual
GeoExT 2023: First International Workshop on Geographic Information Extraction from Texts at ECIR 2023, April 2, 2023,
Dublin, Ireland
∗
Corresponding author.
Envelope-Open niloofar.aflaki@massey.ac.nz (N. Aflaki); kristin.stock@massey.ac.nz (K. Stock); JonesCB@cardiff.ac.uk
(C. B. Jones); h.guesgen@massey.ac.nz (H. Guesgen); y.fukuzawa@massey.ac.nz (Y. Fukuzawa);
jeremy.morley@os.uk (J. Morley)
Orcid 0000-0002-4643-7692 (N. Aflaki); 0000-0002-5828-6430 (K. Stock); 0000-0002-5069-8203 (Y. Fukuzawa)
© 2023 Copyright for this paper by its authors.
CEUR
Workshop
CEUR Workshop Proceedings (CEUR-WS.org)
Proceedings
http://ceur-ws.org
ISSN 1613-0073
1
https://koja.io.ac.nz/
perspectives have also been developed. For example, eImage which focuses on the selection
of an area based on user interests to understand how people feel about different areas [3].
However, web maps that allow users to find the location of places in response to a query rarely
take account of vague and context sensitive terms, and in production systems such as Google
Maps2 and OpenStreetMap3 , places are most commonly found simply by their name, or a type
of feature (e.g., hairdressers). While it is possible to include spatial relation terms (e.g., in, near,
outside, north of) in Google Maps searches, they have little impact on the results returned. There
have been many studies of the use of spatial relational language, e.g., [1, 4, 5, 6, 7, 8, 9], some
of which focus specifically on contextual factors, e.g., [10]. The concept of spatial templates
was introduced in [6] where the template models the applicability or acceptance of locations
relative to a reference object, for a given spatial relation. Only limited progress has been made
in automating the interpretation of spatial language in order to model or predict the locations
referred to in geo-spatial locative expressions e.g. [11, 12, 13], though some progress has also
been made in ’table-top’ space, such as [14] in which spatial templates were modelled with deep
learning. Some notable examples of considering uncertainty in the interpretation of multiple
spatial relationships in locative expressions include [15, 16, 17] which focus on the geometric
characteristics of objects, with limited attention to the geographical types. In the GeoLocate4
service complex locality descriptions are interpreted by determining displacements from a
reference object (relatum) using a mainly rule-based approach, but it does not take context into
account and is limited in the range of spatial relations that can be interpreted. In an early use
of multiple spatial templates to predict geo-spatial location [13] predicted location but did not
consider feature types and only distinguished between urban and rural contexts, while [12]
interpreted relative location descriptions taking account of the contextual factors of place size
and prominence. Logan et al. [6] predicted distance and direction based on spatial relation term
and relatum, using contextual factors but did not consider an explicit locatum. An example of
an experimental search system interface that allows the user to specify one of a small set of
spatial relations is that of [18] in which a search buffer adapts to the size of the reference object,
taking account of predetermined parameters relating to the locatum and relatum types, and
enabling search expansion if no object is found initially. The method adopted here differs from
previous studies in retrieving locations of candidate locata, for a query that specifies locatum
type, spatial relation term selected from a fairly large number of options, and reference object,
based on a regression model that is trained on the combination of spatial relation and context
features.
2. Koja Web Map
As an interactive web map, Koja uses location and language features in a machine learning
model to predict the distance between locatum and relatum for particular spatial relational
terms, in order to select a set of possible features of specified type that lie within a range of the
inferred distance from the relatum. It is intended as a simple prototype for future spatial search
2
https://www.google.com/maps/
3
https://www.openstreetmap.org/
4
https://www.geo-locate.org/
systems that attempt to interpret the meaning of relative spatial relational terminology in a
query. We used a regression model for this prediction, with SMO a regression version of support
vector machine (SVM). The training data contains about 700 location descriptions extracted
from Geograph 5 and Foursquare 6 websites which include locatum, geospatial preposition
and relatum. For a structured query that specifies locatum type, spatial relation and a named
relatum, the derived input features for the regression model include the following:
• GloVe embeddings 7 of the feature types of locatum and relatum.
• The characteristics of the locatum and relatum including geometry type (point, line,
polygon), scale, area, elongation and whether it is liquid or solid. These characteristics
can affect the interpretation of a spatial relation term (e.g., “church beside post office” vs
“road beside river”).
• The density of objects in the area and surrounds.
• The semantics of the spatial relation term based on matching the query spatial relation
with spatial relations in training data. We exploit semantic similarity of spatial relation
terms to support improved learning from training data with similar spatial relations (e.g.,
“beside”, “next to”).
The spatial relation term in combination with the other features are used by the classifier
to infer a distance which will vary for an individual spatial relation depending on the types
of feature (e.g., “park bench near post office” vs “village near London”) [19]. The distance is
converted to a buffer zone to retrieve objects of the specified locatum feature type that lie
inside that zone. Koja supports search based on the specified feature type of the locatum that is
selected from OpenStreetMap types, a choice of 24 different spatial relation terms (e.g. ”near“,
“on”, “at”, “beside”, “opposite”) [20] and, for the reference objects, named places in the City of
London, UK. It is implemented with Leaflet with OpenStreetMap as a base map, and uses models
trained on a set of triples extracted from Geograph
and Foursquare.
The interface allows the user to select the relatum by typing the first three characters of the
name and then selecting from the available places, and then to select the spatial relation term
and locatum type from a drop down list (again, the initial letters can be typed in to narrow
the selection). Once these values are selected and the query submitted, the system determines
the input features and runs them against the model created from training data to predict the
distance that best reflects the specific context of those three terms.
Figure 1 shows the results of the query theatre near Trafalgar Square. The doughnut shape
shows the area that the locatum might be located in, calculated using the distance predicted by
the machine learning model for the expression (in this case, 366.86m), with a margin of error on
either side (based on the mean absolute error divided by average of all distances across all the
expressions in the dataset expressed as a percentage). The orange circles identify the theatres
that fall within the doughnut, and thus that may be considered to fulfil the query theatre near
Trafalgar Square.
5
https://www.geograph.org.uk/
6
https://foursquare.com/
7
https://nlp.stanford.edu/projects/glove/
Figure 1: Predicted locations for theatre near Trafalgar Square
Figure 2: Predicted distance and possible locations for the expression theatre at Trafalgar Square
This distance is different for each selection. Fig 2 shows the results from the query theatre
at Trafalgar Square. As can be seen, the distance that is predicted for the expression using at
(163.66m) is much smaller than for that using near, since the preposition at usually refers to a
closer distance than near.
Figure 3 shows results of the query theatre near St-Martin-in-the-Fields, illustrating the effect
of different context. St-Martin-in-the-Fields is a church, quite close to Trafalgar Square, but
the predicted distance is 149.20m, much smaller than for Trafalgar Square. This is because
Figure 3: Predicted distance and possible locations for the expression theatre near St Martin-in-the-Fields.
the characteristics of the relatum, which is a church (St-Martin-in-the-Fields) differ from the
characteristics of the relatum of Trafalgar Square (in Figure 1), which is a square.
3. Conclusions and Future Work
The Koja web mapping application offers an interactive interface for structured spatial queries
that incorporates context into the interpretation of vague spatial relation terms, providing
distance predictions that use machine learning models with multiple semantic and linguistic
input features. It was trained using examples of georeferenced locative expressions from the
volunteered data sources Geograph and Foursquare. This provides some significant advances
beyond current web mapping applications that either do not cater for spatial relation terms, or
use much simpler models for them.
In future work, the authors plan to work with unstructured queries and to further advance
the models of spatial language by exploiting deep learning approaches, as well as using richer
sets of training data that reflect the geographic context that humans use when formulating
natural language location descriptions.
References
[1] A. Herskovits, Language and spatial cognition, Cambridge university press Cambridge,
1986.
[2] L. Zepner, P. Karrasch, F. Wiemann, L. Bernard, ClimateCharts.net – an interactive
climate analysis web platform, International Journal of Digital Earth 14 (2020) 338–356.
doi:1 0 . 1 0 8 0 / 1 7 5 3 8 9 4 7 . 2 0 2 0 . 1 8 2 9 1 1 2 .
[3] M. S. Barros, A. Degbelo, G. Filomena, Evaluative image 2.0: A web mapping approach to
capture people’s perceptions of a city, Transactions in GIS 26 (2021) 1116–1139. doi:1 0 .
1111/tgis.12867.
[4] K. R. Coventry, S. C. Garrod, Saying, Seeing and Acting, Psychology Press, 2004. doi:1 0 .
4324/9780203641521.
[5] A. Tyler, V. Evans, The semantics of English prepositions: Spatial scenes, embodied
meaning, and cognition, Cambridge University Press, 2003.
[6] S. C. Levinson, LANGUAGE AND SPACE, Annual Review of Anthropology 25 (1996)
353–382. doi:1 0 . 1 1 4 6 / a n n u r e v . a n t h r o . 2 5 . 1 . 3 5 3 .
[7] K. Stock, C. B. Jones, T. Tenbrink, Speaking of location: a review of spatial language
research, Spatial Cognition and Computation 22 (2022) 185–224. doi:1 0 . 1 0 8 0 / 1 3 8 7 5 8 6 8 .
2022.2095275.
[8] P. F. Fisher, T. M. Orf, An investigation of the meaning of near and close on a university
campus, Computers, Environment and Urban Systems 15 (1991) 23–35. doi:1 0 . 1 0 1 6 /
0198- 9715(91)90043- d.
[9] V. Robinson, Individual and multipersonal fuzzy spatial relations acquired using hu-
man–machine interaction, Fuzzy Sets and Systems 113 (2000) 133–145. doi:1 0 . 1 0 1 6 /
s0165- 0114(99)00017- 2.
[10] X. Yao, J.-C. Thill, How far is too far? - a statistical approach to context-contingent
proximity modeling, Transactions in GIS 9 (2005) 157–178. doi:1 0 . 1 1 1 1 / j . 1 4 6 7 - 9 6 7 1 . 2 0 0 5 .
00211.x.
[11] M. M. Hall, C. B. Jones, Generating geographical location descriptions with spatial tem-
plates: a salient toponym driven approach, International Journal of Geographical Informa-
tion Science 36 (2021) 55–85. doi:1 0 . 1 0 8 0 / 1 3 6 5 8 8 1 6 . 2 0 2 1 . 1 9 1 3 4 9 8 .
[12] H. Chen, S. Winter, M. Vasardani, Georeferencing places from collective human descrip-
tions using place graphs, Journal of Spatial Information Science (2018). doi:1 0 . 5 3 1 1 / j o s i s .
2018.17.417.
[13] M. M. Hall, P. D. Smart, C. B. Jones, Interpreting spatial language in image captions,
Cognitive Processing 12 (2010) 67–94. doi:1 0 . 1 0 0 7 / s 1 0 3 3 9 - 0 1 0 - 0 3 8 5 - 5 .
[14] M. Malinowski, M. Fritz, A pooling approach to modelling spatial relations for image
retrieval and annotation, arXiv preprint arXiv:1411.5190 (2014).
[15] Q. Guo, Y. Liu, J. Wieczorek, Georeferencing locality descriptions and computing associ-
ated uncertainty using a probabilistic approach, International Journal of Geographical
Information Science 22 (2008) 1067–1090. doi:1 0 . 1 0 8 0 / 1 3 6 5 8 8 1 0 7 0 1 8 5 1 4 2 0 .
[16] Y. Liu, Q. H. Guo, J. Wieczorek, M. F. Goodchild, Positioning localities based on spatial
assertions, International Journal of Geographical Information Science 23 (2009) 1471–1501.
doi:1 0 . 1 0 8 0 / 1 3 6 5 8 8 1 0 8 0 2 2 4 7 1 1 4 .
[17] P. Doherty, Q. Guo, Y. Liu, J. Wieczorek, J. Doke, Georeferencing incidents from locality
descriptions and its applications: a case study from yosemite national park search and
rescue, Transactions in GIS 15 (2011) 775–793. doi:1 0 . 1 1 1 1 / j . 1 4 6 7 - 9 6 7 1 . 2 0 1 1 . 0 1 2 9 0 . x .
[18] G. Fu, C. B. Jones, A. I. Abdelmoty, Ontology-based spatial query expansion in information
retrieval, in: Lecture Notes in Computer Science, Springer Berlin Heidelberg, 2005, pp.
1466–1482. doi:1 0 . 1 0 0 7 / 1 1 5 7 5 8 0 1 _ 3 3 .
[19] N. Aflaki, K. Stock, C. B. Jones, H. Guesgen, J. Morley, Y. Fukuzawa, What do you mean
you’re in trafalgar square? comparing distance thresholds for geospatial prepositions,
in: 15th International Conference on Spatial Information Theory (COSIT 2022), Schloss
Dagstuhl-Leibniz-Zentrum für Informatik, 2022. doi:1 0 . 4 2 3 0 / L I P I c s . C O S I T . 2 0 2 2 . 1 .
[20] N. Aflaki, K. Stock, C. B. Jones, H. Guesgen, J. Morley, An empirical study of the semantic
similarity of geospatial prepositions and their senses, Spatial Cognition and Computation
23 (2022) 132–176. doi:1 0 . 1 0 8 0 / 1 3 8 7 5 8 6 8 . 2 0 2 2 . 2 1 1 1 6 8 3 .