MatchingLand, geospatial data testbed for the assessment of matching methods

Abstract

This article presents datasets prepared with the aim of helping the evaluation of geospatial matching methods for vector data. These datasets were built up from mapping data produced by official Spanish mapping agencies. The testbed supplied encompasses the three geometry types: point, line and area. Initial datasets were submitted to geometric transformations in order to generate synthetic datasets. These transformations represent factors that might influence the performance of geospatial matching methods, like the morphology of linear or areal features, systematic transformations, and random disturbance over initial data. We call our 11 GiB benchmark data ‘MatchingLand’ and we hope it can be useful for the geographic information science research community.

Assessment of Displacements of Linestrings Based on Homologous Vertexes

“…As data input we have used a set of lines that was obtained from an external independent source. This source dataset is called MatchingLand [29]. The MatchingLand dataset is a testbed composed of geographic features (points, lines and areas) whose content is intended as an experimental design, as will be seen below.…”

“…A factor is the independent variable (e.g., morphology, positional error, road category, etc.) for which we want to determine its effect over the controlled variable, and the different values of this factor are named levels (e.g., for the morphology, the five classes presented in Xavier et al [29]. Treatment is the combination of different levels of factors considered in an experiment (e.g., the combination of morphological classes and different positional error cases).…”

“…The three cases of random perturbations (5, 12.5, and 25 m) are coded in Tables 2 and 3 as ra1, ra2, and ra3, respectively. MatchingLand obtained these perturbed cases using a new method of vector fields that was based on three parameters (standard displacement, field resolution, and sigma) [29].…”

“…Non-perturbed random, translation, rotations, and scaling cases are coded as ra0, t0, r0, and s0, respectively, in Tables 2 and 3. More information about the obtaining of the perturbed lines is available in [29].…”

Optimal spatial data matching for conflation: A network flow‐based approach

“…It should be noted that other test datasets can generally be used for testing conflation methods. One example is the dataset of Xavier, Ariza‐López, and Ureña‐Cámara (), built up from maps by official Spanish mapping agencies.…”