Created
March 21, 2025 00:14
-
-
Save kaligrafy/97280e37fb7b3556eda1a8171e19b1f9 to your computer and use it in GitHub Desktop.
ISCTSC 2025 POIs Detailed Literature Review
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| # Literature Review | |
| ## Study on the spatial heterogeneity of the POI quality in OpenStreetMap | |
| ### bibtex | |
| @inproceedings{Yang2018, | |
| title={Study on the spatial heterogeneity of the POI quality in OpenStreetMap}, | |
| author={Shuai Yang and Jie Shen and Milan Kone{\vc}n{\'y} and Yong Wang and Radim {\vS}tampach}, | |
| year={2018}, | |
| url={https://api.semanticscholar.org/CorpusID:108396134} | |
| } | |
| ### Location | |
| - Beijing and Xian in China | |
| ### Method of position accuracy assessment | |
| - automated/grid-based | |
| - OSM compared with reference: Baidu Maps | |
| - euclidean distance | |
| - topological relationship between the POIs and the road networks | |
| ### Summary keywords | |
| - focus on POIs data in OSM, comparing and analyzing the data quality of POIs, and to find out the spatial heterogeneity of the data quality of POIs in different regions | |
| - explore the relationship between local characteristics and the quality data of POIs in OSM in different regions | |
| - used geographically weighted regression (GWR) to assess positional accuracy, data completeness and topological consistency | |
| - VGI: no strict data quality control/data quality uncertain | |
| - VGI: malicious/vandalism | |
| - Scope: comparing and analyzing OSM POI data quality | |
| - methods: positional accuracy, attribute accuray, currency, completeness, logical consistency, lineage | |
| - compare with a reference (ground-truth!?) dataset using positional accuracy, completeness and topological consistency (after data matching) | |
| - compare using OSM data history | |
| - assess spatial heterogeneity | |
| - analysing the relationship between the data quality of POIs in OSM and local characteristics: population density, economic status, distribution of volunteers/OSM editors | |
| - explanation of OSm data (node/way/relation) | |
| ## OVERTURE POI DATA FOR THE UNITED KINGDOM: A COMPREHENSIVE, QUERYABLE OPEN DATA PRODUCT | |
| ### bibtex | |
| @article{doi:10.1177/23998083241263124, | |
| author = {Patrick Ballantyne and Cillian Berragan}, | |
| title ={Overture Point of Interest data for the United Kingdom: A comprehensive, queryable open data product, validated against Geolytix supermarket data}, | |
| journal = {Environment and Planning B: Urban Analytics and City Science}, | |
| volume = {51}, | |
| number = {8}, | |
| pages = {1974-1980}, | |
| year = {2024}, | |
| doi = {10.1177/23998083241263124}, | |
| URL = {https://doi.org/10.1177/23998083241263124}, | |
| eprint = { https://doi.org/10.1177/23998083241263124} | |
| } | |
| ### Location | |
| - UK | |
| ### Method of position accuracy assessment | |
| - distance between reference and Overture data | |
| ### Summary keywords | |
| - Overture Maps data | |
| - Overture data format | |
| - difficulty to export data for non-technical users | |
| - compared Overture with Geolytix Supermarket Retail Points (GSRP) | |
| - checked median distance between Overture and GSRP | |
| - Problems with attribute completeness in Overture data | |
| - no ground truth data or manual validation | |
| ## On-line Aggregation of POIs from Google and Facebook | |
| ### bibtex | |
| @inproceedings{10.1145/3297280.3297576, | |
| author = {Toccu, Maurizio and Psaila, Giuseppe and Altomare, Davide}, | |
| title = {On-line aggregation of POIs from Google and Facebook}, | |
| year = {2019}, | |
| isbn = {9781450359337}, | |
| publisher = {Association for Computing Machinery}, | |
| address = {New York, NY, USA}, | |
| url = {https://doi.org/10.1145/3297280.3297576}, | |
| doi = {10.1145/3297280.3297576}, | |
| booktitle = {Proceedings of the 34th ACM/SIGAPP Symposium on Applied Computing}, | |
| pages = {1081–1089}, | |
| numpages = {9}, | |
| keywords = {Google and Facebook, aggregating place information, geo-located places and POI, geographic information retrieval, social media}, | |
| location = {Limassol, Cyprus}, | |
| series = {SAC '19} | |
| } | |
| ### Location | |
| - ~300 randomly selected public spaces in Manchester (UK) | |
| ### Method of position accuracy assessment | |
| - distance between Google POIs and Facebook Pages with locations | |
| ### Summary keywords | |
| - poi aggregation | |
| - compare names, geolocation and addresses to match between Google POIs and Facebook POIs | |
| - extreme volatility of social media (changes in attributes are continuous) | |
| - considers the linear relationship between distance and membership degree (trust that this is the same location): 0-100m: 100%, 2km: 0%, linear in-between | |
| - string/names matching: Levenstein distance, phonetic Distance | |
| ## Comparison of the accuracy of OpenStreetMap for Ireland with Google Maps and Bing Maps | |
| ### bibtex | |
| @article{article, | |
| author = {Cipeluch, Blazej and Jacob, Ricky and Winstanley, A. and Mooney, Peter}, | |
| year = {2010}, | |
| month = {01}, | |
| pages = {}, | |
| title = {Comparison of the accuracy of OpenStreetMap for Ireland with Google Maps and Bing Maps} | |
| } | |
| ### Location | |
| - Ireland | |
| ### Method of position accuracy assessment | |
| - distance between Google/Bing and OSM and count of POIs | |
| ### Summary keywords | |
| - Compared road networks between OSM, Bing and Google Maps | |
| - Compared POIs but considered that OSM was the ground-truth for geolocation (?!) | |
| ## A kilometer or a mile? Does buffer size matter when it comes to car ownership? | |
| ### bibtex | |
| @article{LAVIOLETTE2022103456, | |
| title = {A kilometer or a mile? Does buffer size matter when it comes to car ownership?}, | |
| journal = {Journal of Transport Geography}, | |
| volume = {104}, | |
| pages = {103456}, | |
| year = {2022}, | |
| issn = {0966-6923}, | |
| doi = {https://doi.org/10.1016/j.jtrangeo.2022.103456}, | |
| url = {https://www.sciencedirect.com/science/article/pii/S096669232200179X}, | |
| author = {Jérôme Laviolette and Catherine Morency and E.O.D. Waygood}, | |
| keywords = {Car ownership, Accessibility, Built environment, Machine learning, Gradient boosting machines, Distance thresholds} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - Is there a threshold value beyond or below which there is no more | |
| measurable impact on household decisions? | |
| ## When local access matters: A detailed analysis of place, neighbourhood amenities and travel choice | |
| ### bibtex | |
| @article{doi:10.1177/0042098020951001, | |
| author = {Erik Elldér and Katarina Haugen and Bertil Vilhelmson}, | |
| title ={When local access matters: A detailed analysis of place, neighbourhood amenities and travel choice}, | |
| journal = {Urban Studies}, | |
| volume = {59}, | |
| number = {1}, | |
| pages = {120-139}, | |
| year = {2022}, | |
| doi = {10.1177/0042098020951001}, | |
| URL = {https://doi.org/10.1177/0042098020951001}, | |
| eprint = {https://doi.org/10.1177/0042098020951001} | |
| } | |
| ### Location | |
| - Västra Götaland, Sweden (Swedish national travel survey) | |
| - GILDA Sweden databse: 100m resolution for amenities (is it the building centroid, the main entrance, the parking lot, the whole land?) | |
| ### Method of position accuracy assessment | |
| - None, 1km radius for amenities accessibility | |
| ### Summary keywords | |
| - Rather than deriving from such aggregate measures, travel to many everyday activities is likely directly influenced by the presence of particular neighbourhood amenities (e.g. shops, restaurants and schools) that enable locally oriented daily living | |
| - Therefore, from the perspectives of both travel behaviour theory and sustainable urban development policy, closer investigation of the distinct role of local accessibility in terms of the number, types and variety of amenities available in urban neighbourhoods is justified | |
| - analysis of the modal shares related to the number of amenities in a 1km radius around home (is it network distance or as the crow flies/euclidian?), and no weighting of the amenities by attractivness or building/amenity floor area or any other attribute. | |
| ## Assessing Crowdsourced POI Quality: Combining Methods Based on Reference Data, History, and Spatial Relations | |
| ### bibtex | |
| @Article{ijgi6030080, | |
| AUTHOR = {Touya, Guillaume and Antoniou, Vyron and Olteanu-Raimond, Ana-Maria and Van Damme, Marie-Dominique}, | |
| TITLE = {Assessing Crowdsourced POI Quality: Combining Methods Based on Reference Data, History, and Spatial Relations}, | |
| JOURNAL = {ISPRS International Journal of Geo-Information}, | |
| VOLUME = {6}, | |
| YEAR = {2017}, | |
| NUMBER = {3}, | |
| ARTICLE-NUMBER = {80}, | |
| URL = {https://www.mdpi.com/2220-9964/6/3/80}, | |
| ISSN = {2220-9964}, | |
| DOI = {10.3390/ijgi6030080} | |
| } | |
| ### Location | |
| - Paris region, France | |
| ### Method of position accuracy assessment | |
| - compare with IGN database (official government provided data) | |
| - geotagged Flickr images | |
| - distance between centroid of all historical position and the latest position | |
| - distance to the building center | |
| - distance to the nearest building boundary, as a proxy for amenity entrance | |
| - visual inspection (partial, not explicit) | |
| ### Summary keywords | |
| - the consistency of the POI locations is very important to reduce the map reader cognitive load. | |
| - it is not uncommon for VGI to suffer from participation biases at all levels of granularity. | |
| - VGI: existing quality evaluation methods do not always provide the means to evaluate them. | |
| - VGI has matured enough to be considered as a replacement of, or as a way to enrich, authoritative data with GI, which has, so far, been missing from authoritative databases. | |
| - Recent research has proposed methods to evaluate VGI POIs based on history or on comparisons with reference datasets, but neither is enough to assess quality alone. | |
| - examined the quality of OSM points against authoritative ones from IGN (The French Mapping Agency) by using a data matching algorithm (did you validate the accuracy of IGN data beforehand?) | |
| - geotagged Flickr images have been used to validate the accuracy of OSM points + satellite imagery (for all POIs?, not explained) | |
| - estimated the accuracy using OSM history (number of changes, displacements/distances of moved POIs, etc.): distance between centroid of all historical position and the latest position | |
| - only named POIs analyzed, which is a problem because some POIs are correctly categorize (bakery, shop, etc.) but have no name added in OSM, but should be analyzed/validated anyway | |
| - the authors says that sometimes, a visual inspection would be needed, but they did not do it explicitly | |
| - they checked three methods to get the POI "correct" location: inside a building, distance to the building center, distance to the nearest building boundary, as a proxy for amenity entrance | |
| - no check if the POIs are indeed mapped in or near the correct building | |
| - great analysis of how the things are mapped, POIs inside buildings, school area vs their multiple buildings, bus stops, ATM, subway stations with their multiple entrances, etc. | |
| - VGI projects, such as OSM, can be more accurate and more complete than authoritative data (especially in developing countries), which violates the basic assumption of using authoritative data as reference data | |
| ## Point-of-Interest (POI) Data Validation Methods: An Urban Case Study | |
| ### bibtex | |
| @Article{ijgi10110735, | |
| AUTHOR = {Yeow, Lih Wei and Low, Raymond and Tan, Yu Xiang and Cheah, Lynette}, | |
| TITLE = {Point-of-Interest (POI) Data Validation Methods: An Urban Case Study}, | |
| JOURNAL = {ISPRS International Journal of Geo-Information}, | |
| VOLUME = {10}, | |
| YEAR = {2021}, | |
| NUMBER = {11}, | |
| ARTICLE-NUMBER = {735}, | |
| URL = {https://www.mdpi.com/2220-9964/10/11/735}, | |
| ISSN = {2220-9964}, | |
| DOI = {10.3390/ijgi10110735} | |
| } | |
| ### Location | |
| - Tampines town, Singapore | |
| ### Method of position accuracy assessment | |
| - distance between POIs in different datasets (euclidian distance) | |
| - spatial analysis tools (Nearest neighbour, k-nearest neighbours, spatial autocorrelation, spatial clustering, etc.) | |
| - name match using Levenshtein distance, Longest Common Subsequence (LCS), the mean of the Token Sort Ratio and Token Set Ratio, WordNet Similarity Metric, etc. | |
| ### Summary keywords | |
| - OSM, Google Maps, HERE Maps, OneMap (authoritative national map of Singapore) and Singapore Land Authority (SLA) POI dataset used as reference data | |
| - challenging to compare between different POI sources as they do not follow a standard set of validation metrics that comprehensively and objectively evaluate different aspects of their databases’ data quality | |
| - great review of approaches for validating POI data quality with number of related articles per year from 2010 to 2021, plus methods comparison table (table 1) | |
| - explained completeness, logical consistency, positional accuracy, temporal quality, thematic accuracy (name and place type), usability/fitness-for-use (ISO 19157: Geographic Information: Data Quality) | |
| - discussed map scale resolution and zoom level to assess positional accuracy | |
| - evaluation methods using building/area footprints are exluded because not often available | |
| - ~ 50% of POIs in OSM/Google/HERE are in the reference datasets | |
| - median distances between OSM/Google/HERE and reference datasets are between 8 and 25m, with third quartile going up to almost 70m | |
| ## Defining Fitness-for-Use for Crowdsourced Points of Interest (POI) | |
| ### bibtex | |
| @Article{ijgi5090149, | |
| AUTHOR = {Jonietz, David and Zipf, Alexander}, | |
| TITLE = {Defining Fitness-for-Use for Crowdsourced Points of Interest (POI)}, | |
| JOURNAL = {ISPRS International Journal of Geo-Information}, | |
| VOLUME = {5}, | |
| YEAR = {2016}, | |
| NUMBER = {9}, | |
| ARTICLE-NUMBER = {149}, | |
| URL = {https://www.mdpi.com/2220-9964/5/9/149}, | |
| ISSN = {2220-9964}, | |
| DOI = {10.3390/ijgi5090149} | |
| } | |
| ### Location | |
| - only examples | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - assist users of a POI dataset in choosing from the range of available quality measures and assessment methods those which are appropriate for evaluating its fitness-for-use with regards to their specific use case or application task | |
| - uses ISO assessment methods (completeness, logical consistency, positional accuracy, temporal quality, thematic accuracy, and usability) | |
| - 3 main attributes of POI: name, location (lat/lon) and category | |
| - different level of accuracy acceptability for different use cases (small restaurant or cafe vs large airport or shopping mall) or when name and/or category is available or not: example: an exact location with only a categroy like restaurant can be found, while an approximate location with a name like "restaurant" cannot be found if multiple restaurant are found in the vicinity | |
| - the traditional assumption of authoritative or commercial datasets being of a higher quality compared to VGI is no longer fully reliable. | |
| ## Assessing the quality of OpenStreetMap contributors together with their contributions | |
| ### bibtex | |
| @inproceedings{inproceedings, | |
| author = {Jokar Arsanjani, Jamal and Barron, Christopher and Bakillah, Mohamed and Helbich, Marco}, | |
| year = {2013}, | |
| month = {01}, | |
| pages = {}, | |
| title = {Assessing the Quality of OpenStreetMap Contributors together with their Contributions} | |
| } | |
| ### Location | |
| - Heidelberg, Germany | |
| ### Method of position accuracy assessment | |
| - compared with reference road network | |
| - by type of OSM user/contributor | |
| ### Summary keywords | |
| - analyzed road network quality, not POI quality | |
| - at the beginning of OSM, five major groups of users are characterized: neophytes, interested amateurs, expert amateurs, expert professionals, and expert authorities. | |
| - According to Zipf, OSM mappers into several categorizes based on the quantity of contributions as hit-and-run mappers, newbies, casual mappers, heavy mappers, heavy mappers 2.0, addicted mappers, crazy mappers, and bots | |
| - cross comparing contributors' data with their quality | |
| - OSM dataset compared with data from Federal Agency for Cartography and Geodesy (BKG) | |
| ## Quality Assessment of OpenStreetMap’s Points of Interest with Large-Scale Real Data | |
| ### bibtex | |
| @article{doi:10.1177/03611981231169280, | |
| author = {Christian Klinkhardt and Fabian Kühnel and Michael Heilig and Sven Lautenbach and Tim Wörle and Peter Vortisch and Tobias Kuhnimhof}, | |
| title ={Quality Assessment of OpenStreetMap’s Points of Interest with Large-Scale Real Data}, | |
| journal = {Transportation Research Record}, | |
| volume = {2677}, | |
| number = {12}, | |
| pages = {661-674}, | |
| year = {2023}, | |
| doi = {10.1177/03611981231169280}, | |
| URL = {https://doi.org/10.1177/03611981231169280}, | |
| eprint = {https://doi.org/10.1177/03611981231169280} | |
| } | |
| ### Location | |
| - 49 areas of ~350m radius, mostly in 17 large cities in Germany but also in smaller cities, but all in urban regions (total of 1305 POIs for private businesses, 379 shops) | |
| ### Method of position accuracy assessment | |
| - ground-truth, surveyed POI data (yes!) | |
| ### Summary keywords | |
| - we need POI accuracy and precision for travel demand modeling | |
| - hard to use proprietary datasets for travel demand modeling because the methodology is not well known | |
| - proprietary datasets are costly | |
| - surveyed 49 real-world areas and surveyed in-place POI data (took between 20 min and 3 h to complete survey) | |
| - not enough POI useage in transport modelling (mostly use road/transit network data and or density or landuse) | |
| - For the purpose of travel demand estimation, object completeness and correct classification of POI are of high importance while for navigation, correctness of topological relations plays as important role as well | |
| - Quality assessment of POI is already a popular subject in the field of geo-information. The approaches still barely focus on the completeness of POI and, if so, they focus on intrinsic assessments or the comparison of dif- ferent sources of VGI. Completeness comparison with "ground truth data" is found only in a few cases and in geographically and POI type limited applications. Furthermore, data quality is highly dependent on the local community, leading to different results in quality assessment. A large-scale quality assessment with real- world data can therefore enable further insights into the usability of VGI data. | |
| - 22% of private business were in OSM | |
| - 73% of shops were in OSM | |
| ## OSM POI ANALYZER: A PLATFORM FOR ASSESSING POSITION OF POIs IN OPENSTREETMAP | |
| ### bibtex | |
| @Article{isprs-archives-XLII-2-W7-497-2017, | |
| AUTHOR = {Kashian, A. and Rajabifard, A. and Chen, Y. and Richter, K. F.}, | |
| TITLE = {OSM POI ANALYZER: A PLATFORM FOR ASSESSING POSITION OF POIs IN OPENSTREETMAP}, | |
| JOURNAL = {The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences}, | |
| VOLUME = {XLII-2/W7}, | |
| YEAR = {2017}, | |
| PAGES = {497--504}, | |
| URL = {https://isprs-archives.copernicus.org/articles/XLII-2-W7/497/2017/}, | |
| DOI = {10.5194/isprs-archives-XLII-2-W7-497-2017} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - Normally, OSM data quality can be assessed by two different methods: (1) using reference or authoritative resources to compare with existing OSM data; (2) establishing rules (manual or automatic) and checking incoming data with these rules to detect mistakes automatically (like osmose, keep right, maproulette, etc.). | |
| - the authors proposes a tool to analyze POIs in OSM | |
| - checks that for example a gas station is near a road, or a ferry terminal close to water, etc. | |
| - the tool is no longer maintained/available | |
| ## Undergraduate Research in Action: Evaluating the positional differences between the Google Maps and the United States Census Bureau geocoding APIs | |
| ### bibtex | |
| @unknown{unknown, | |
| author = {Bucklew, Matthew and Curry, Caitlin and Krach, Noah and Molnar, Jessica and Stancil, Robert and Tilghman, John and Young, Josh and Lembo, Arthur}, | |
| year = {2016}, | |
| month = {10}, | |
| pages = {}, | |
| title = {Undergraduate Research in Action: Evaluating the positional differences between the Google Maps and the United States Census Bureau geocoding APIs}, | |
| doi = {10.13140/RG.2.2.10457.93287} | |
| } | |
| ### Location | |
| - 106 primarily residentialknown addresses geolocated with Google API and the US Census bureau API, United States | |
| ### Method of position accuracy assessment | |
| - ground-truth = building centroids, mostly detached houses | |
| ### Summary keywords | |
| - 96% supplied to Google API were successful, with an average distance to real location of 81m | |
| - 84% supplied to US Census bureau API were successful, with an average distance to real location of 271m | |
| ## Points of Interest (POI): a commentary on the state of the art challenges, and prospects for the future | |
| ### bibtex | |
| @Article{Psyllidis2022, | |
| author={Psyllidis, Achilleas and Gao, Song and Hu, Yingjie and Kim, Eun-Kyeong and McKenzie, Grant and Purves, Ross and Yuan, May and Andris, Clio}, | |
| title={Points of Interest (POI): a commentary on the state of the art, challenges, and prospects for the future}, | |
| journal={Computational Urban Science}, | |
| year={2022}, | |
| month={Jun}, | |
| day={28}, | |
| volume={2}, | |
| number={1}, | |
| pages={20}, | |
| issn={2730-6852}, | |
| doi={10.1007/s43762-022-00047-w}, | |
| url={https://doi.org/10.1007/s43762-022-00047-w} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - talks about POIs in general, their usage, why they exists, their interactions, their longevity, who uses them, etc. | |
| - lists POI data sources: Big-tech companies, such as Yelp, Foursquare, Google Places, and Facebook, generate rich amounts of POI data, and similar prominent companies include Baidu and Gaode Maps (in China). OpenStreetMap (OSM) | |
| - Additional sources include Niche.com, Wiki- mapia, TomTom, and HERE map, OneMap, or POIs that are provided by mobility data companies such as SafeGraph and Cuebiq. | |
| - explain problems when POIs are only represented as points, like parks or places with multiple entrances (especially for disabled people) | |
| - discuss temporality (hours of operation, closures, changes, moving, etc.) | |
| - usually, commercial POI providers will tell you when a POI was closed (temporarly or permanently), but will not tell you if the POI is new or its age, while OSM can provide such info when appropriate tags are filled. | |
| - usage of street-view and web data using deep learning to update/enhance POI data | |
| - POI data is used for a lot of things, including public health, urban planning, sustainable development, mobility, community studies, and sociology. | |
| - sometimes, there are erros in categorization | |
| - accessibility analysis, with precise access information is often missing | |
| ## A Comprehensive Framework for Intrinsic OpenStreetMap Quality Analysis | |
| ### bibtex | |
| @article{https://doi.org/10.1111/tgis.12073, | |
| author = {Barron, Christopher and Neis, Pascal and Zipf, Alexander}, | |
| title = {A Comprehensive Framework for Intrinsic OpenStreetMap Quality Analysis}, | |
| journal = {Transactions in GIS}, | |
| volume = {18}, | |
| number = {6}, | |
| pages = {877-895}, | |
| doi = {https://doi.org/10.1111/tgis.12073}, | |
| url = {https://onlinelibrary.wiley.com/doi/abs/10.1111/tgis.12073}, | |
| eprint = {https://onlinelibrary.wiley.com/doi/pdf/10.1111/tgis.12073}, | |
| year = {2014} | |
| } | |
| ### Location | |
| - Madrid, Spain | |
| - San Francisco, USA | |
| - Yaoundé, Cameroon | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - analyzed OSM data (road network and addresses, nothing specifically on POIs) | |
| - ground truth reference datasets: but we should say this is not the case most of the time... | |
| - developed iOSMAnalyzer framework (cannot find it online, old: 2013) | |
| - fitness for purpose categories: | |
| - General Information on the Study Area | |
| - Routing and Navigation | |
| - Geocoding (needs address) | |
| - Points of Interest-Search | |
| - Map- Applications | |
| - User Information and Behavior | |
| - 25 different intrinsic quality indicators used in the framework | |
| ## Citizens as sensors: the world of volunteered geography | |
| ### bibtex | |
| @Article{Goodchild2007, | |
| author={Goodchild, Michael F.}, | |
| title={Citizens as sensors: the world of volunteered geography}, | |
| journal={GeoJournal}, | |
| year={2007}, | |
| month={Aug}, | |
| day={01}, | |
| volume={69}, | |
| number={4}, | |
| pages={211-221}, | |
| issn={1572-9893}, | |
| doi={10.1007/s10708-007-9111-y}, | |
| url={https://doi.org/10.1007/s10708-007-9111-y} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - coined the term volunteered geographic information (VGI) | |
| - humans as sensors | |
| - citizen science: describe communities or networks of citizens who act as observers in some domain of science. They need to be trianed or self-trained and have expertise in the domain involved. | |
| - discuss issues with alignements, DATUM, etc. | |
| ## Fine-resolution population mapping using OpenStreetMap points-of-interest | |
| ### bibtex | |
| @article{Bakillah02092014, | |
| author = {Mohamed Bakillah, Steve Liang, Amin Mobasheri, Jamal Jokar Arsanjani and Alexander Zipf}, | |
| title = {Fine-resolution population mapping using OpenStreetMap points-of-interest}, | |
| journal = {International Journal of Geographical Information Science}, | |
| volume = {28}, | |
| number = {9}, | |
| pages = {1940--1963}, | |
| year = {2014}, | |
| publisher = {Taylor \& Francis}, | |
| doi = {10.1080/13658816.2014.909045}, | |
| URL = {https://doi.org/10.1080/13658816.2014.909045}, | |
| eprint = {https://doi.org/10.1080/13658816.2014.909045} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - explains how to apply ISO data quality elements to VGI, especially OSM data. | |
| - indicators to assess vgi quality: data indicators,demographic indicators, socio-economic indicators, contributors' indicators | |
| - compare VGI against authoritative datasets | |
| - research far from complete on the subject | |
| ## Crowdsourced geospatial data quality: challenges and future directions | |
| ### bibtex | |
| @article{article, | |
| author = {Basiri, Anahid and Haklay, Muki and Foody, Giles and Mooney, Peter}, | |
| year = {2019}, | |
| month = {05}, | |
| pages = {1-6}, | |
| title = {Crowdsourced geospatial data quality: challenges and future directions}, | |
| volume = {33}, | |
| journal = {International Journal of Geographical Information Science}, | |
| doi = {10.1080/13658816.2019.1593422} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - talks about VGI biases | |
| - the fact that contributors are volunteering is a bias in itself: not necessarily representative of the population | |
| ## On predicting and improving the quality of Volunteer Geographic Information projects | |
| ### bibtex | |
| @article{Bordogna01022016, | |
| author = {Gloria Bordogna, Paola Carrara, Laura Criscuolo, Monica Pepe and Anna Rampini}, | |
| title = {On predicting and improving the quality of Volunteer Geographic Information projects}, | |
| journal = {International Journal of Digital Earth}, | |
| volume = {9}, | |
| number = {2}, | |
| pages = {134--155}, | |
| year = {2016}, | |
| publisher = {Taylor \& Francis}, | |
| doi = {10.1080/17538947.2014.976774}, | |
| URL = {https://doi.org/10.1080/17538947.2014.976774}, | |
| eprint = {https://doi.org/10.1080/17538947.2014.976774} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - talks mainly about the use of VGI in scientific contexts | |
| - gives examples of VGI projects and relations with authorities | |
| - many researchers are skeptical about the quality of VGI for use in scientific domain and academia | |
| - quality depends on a lot of factors | |
| - quality in VGI projects/Citizen Science is based on: | |
| - type of the contributions requested to the volunteers | |
| - type of task | |
| - characteristics of the agents/volunteers | |
| - technologies used to manage/edit the data | |
| - types of volunteers: | |
| - neophyte | |
| - interested amateur | |
| - expert amateur | |
| - expert authority | |
| - unaware volunteer | |
| - ways of VGI creation: | |
| - automatic and implicit | |
| - manual and implicit | |
| - manual and explicit (volunteer is aware) | |
| - automatic and explicit | |
| - mixed strategy | |
| - summaries methods to assess volunteers produced data quality | |
| - ex ante (before editing) and ex post (after editing) approaches to improve quality | |
| ## Assessing shop completeness in OpenStreetMap for two federal states in Germany | |
| ### bibtex | |
| @Article{agile-giss-2-20-2021, | |
| AUTHOR = {Brückner, J. and Schott, M. and Zipf, A. and Lautenbach, S.}, | |
| TITLE = {Assessing shop completeness in OpenStreetMap for two federal states in Germany}, | |
| JOURNAL = {AGILE: GIScience Series}, | |
| VOLUME = {2}, | |
| YEAR = {2021}, | |
| PAGES = {20}, | |
| URL = {https://agile-giss.copernicus.org/articles/2/20/2021/}, | |
| DOI = {10.5194/agile-giss-2-20-2021} | |
| } | |
| ### Location | |
| - Baden-Württemberg, Germany | |
| - Saxony, Germany | |
| ### Method of position accuracy assessment | |
| - used a saturation parameter based on history of OSM data (cumulative count of shops in the region over time): if the growth rate plateau to 0, it is considered saturated and 100% complete (?!) | |
| - completeness ranged between 42% and 100% depending on the region, while the majority achieved at least 80% (average of 86%) | |
| - no ground-truth validation | |
| - no statistically significant difference in completeness between rural and urban regions, which is counter-intuitive | |
| - regions of low unemployment rate had higher completeness | |
| ### Summary keywords | |
| - analyzed rural and urban districts | |
| ## A taxonomy of quality assessment methods for volunteered and crowdsourced geographic information | |
| ### bibtex | |
| @article{article, | |
| author = {Degrossi, Livia and De Albuquerque, Joao and Rocha, Roberto and Zipf, Alexander}, | |
| year = {2018}, | |
| month = {04}, | |
| pages = {}, | |
| title = {A taxonomy of quality assessment methods for volunteered and crowdsourced geographic information}, | |
| volume = {22}, | |
| journal = {Transactions in GIS}, | |
| doi = {10.1111/tgis.12329} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - coined the term CGI (Crowdsourced Geospatial Information) | |
| - CGI includes VGI, but also data collected without the will or conscious knowledge of the provider | |
| - problem: previous studies do not differentiate clearly between assessment methods that require a reference dataset and methods that can be used when no comparable data is available. The latter is probably the most frequent situation in practice, since the use of CGI is often motivated by a lack of availability (or currency) of authoritative data sources, so "What types of methods can be employed to assess the quality of CGI in the absence of authoritative data?" | |
| - three types of CGI: | |
| - social media | |
| - crowd sensing | |
| - collaborative mapping | |
| - methods/approaches of assessing quality: | |
| - crowdsourcing approach (ability on a group of people to validate and correct the data) | |
| - social approach (also called hierarchical approach: relies on hierarchy of individuals who act as moderators or gatekeepers) | |
| - geographic approach (based on data comparison between geographic information and the body of geographic knowledge) | |
| - two types of quality assessment | |
| - extrinsic (comparison with authoritative data) | |
| - intrinsic (analyze historical metadata, user behavior, etc.) | |
| - ex ante and ex post approaches to improve quality | |
| - development of a taxonomy: | |
| 1. Iteration 1: | |
| - Reference dimension: | |
| - Extrinsic | |
| - Intrinsic | |
| 2. Iteration 2: | |
| - Object dimension: | |
| - Content | |
| - Volunteer | |
| - Approach dimension: | |
| - Crowdsourcing | |
| - Hierarchical | |
| - Geographic | |
| 3. Iteration 3: | |
| - Temporal dimension: | |
| - Ex post | |
| - All (ex ante and ex post) | |
| 4. Iteration 4: | |
| - Criteria dimension: | |
| - Positional accuracy | |
| - Thematic accuracy | |
| - Fitness-for-use | |
| - Trust | |
| - Reliability | |
| - Plausibility | |
| 5. Iteration 5: | |
| - There are not any new characteristics and dimensions that can be obtained from the methods | |
| - Great proposition for systhematic review of CGI data | |
| ## Point of Interest Matching between Different Geospatial Datasets | |
| ### bibtex | |
| @article{article, | |
| author = {Deng, Hongzhong and Luo, Chao and Liu, and Wang, Cheng}, | |
| year = {2019}, | |
| month = {10}, | |
| pages = {435}, | |
| title = {Point of Interest Matching between Different Geospatial Datasets}, | |
| volume = {8}, | |
| journal = {ISPRS International Journal of Geo-Information}, | |
| doi = {10.3390/ijgi8100435} | |
| } | |
| ### Location | |
| - China (with Gaode Map: 2.7M POIs and Baidu Map: 1.9M POIs, but randomly selected 392 POIs in Baidu which were matched with 323 POIs in Gaode) | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - two steps for matching: | |
| - matching the POIs between databases (most difficult, but most crucial) | |
| - combining attributes | |
| - multiattribute matching method based on the Dempster–Shafer (D–S) evidence theory | |
| - problems with text matching: A large number of spatial entity matching methods use traditional text similarity calculation methods, such as the Levenshtein distance. The space-time efficiency is low, and the semantic similarity between text attributes cannot be accurately quantified | |
| - how to deal with attributes conflicts? | |
| - method proposed based on D-S evidence theory | |
| - compared Gaode Map and Baidu Map (China only) | |
| - results: | |
| - name is better than address for matching (there could be more than one POI at an address) | |
| - problems: | |
| - only applicable with datasets using a perfect category system | |
| - must be re-implemented/re-calibrated in each language | |
| ## DuMapper: Towards Automatic Verification of Large-Scale POIs with Street Views at Baidu Maps | |
| ### bibtex | |
| @article{article, | |
| author = {Fan, Miao and Huang, Jizhou and Wang, Haifeng}, | |
| year = {2024}, | |
| month = {11}, | |
| pages = {}, | |
| title = {DuMapper: Towards Automatic Verification of Large-Scale POIs with Street Views at Baidu Maps}, | |
| doi = {10.48550/arXiv.2411.18073} | |
| } | |
| ### Location | |
| - China (Baidu Maps) | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - used Street View imagery to verify POIs | |
| - created DuMapper, an automatic system for verifying POIs using street views from Baidu Maps | |
| - used over at least 3.5 years over 405 millions iterations of POI verification | |
| - proposing a new version (II) with deep multimodal embedding (using a convolutional neural net: CNN) and approximate nearest neighbor (ANN) search | |
| - some issues with the precision of photo geo-location and the performance needed to analyse billions of images to find POIs signs | |
| - results after training with train, valid and test datasets with accuracy validation: | |
| - DuMapper I: between 75 and 85% (between 1.7 and 3 queries per second) | |
| - DuMapper II: between 85 and 91% (between 7.4 and 152 queries per second) | |
| - Expert Mapper: 95% (0.007 queries per second) | |
| - great: source code publically available | |
| ## Integration of authoritative and volunteered geographic information for updating urban mapping: challenges and potentials | |
| ### bibtex | |
| @article{article, | |
| author = {Fernandes, Vivian and Elias, Elias and Zipf, Alexander}, | |
| year = {2020}, | |
| month = {08}, | |
| pages = {}, | |
| title = {INTEGRATION OF AUTHORITATIVE AND VOLUNTEERED GEOGRAPHIC INFORMATION FOR UPDATING URBAN MAPPING: CHALLENGES AND POTENTIALS}, | |
| volume = {XLIII-B4-2020}, | |
| journal = {ISPRS - International Archives of the Photogrammetry Remote Sensing and Spatial Information Sciences}, | |
| doi = {10.5194/isprs-archives-XLIII-B4-2020-261-2020} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - literature review and analysis of paper provenance | |
| - review of most attempst to integrate authoritative and VGI, by country or research group | |
| ## Assessing VGI Data Quality | |
| ### bibtex | |
| @inbook{inbook, | |
| author = {Fonte, Cidalia and Antoniou, Vyron and Bastin, Lucy and Estima, Jacinto and Jokar Arsanjani, Jamal and Laso Bayas, Juan and See, Linda and Vatseva, Rumiana}, | |
| year = {2017}, | |
| month = {09}, | |
| pages = {137-163}, | |
| title = {Assessing VGI Data Quality}, | |
| isbn = {EPUB 978-1-911529-18-7}, | |
| doi = {10.5334/bbf.g} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - lack of rigorous data specifications in VGI compared to authoritative data | |
| - frequent changes are good, but can deteriorate the quality of the data over time | |
| - grid-based or small-area based organized editing can be good but can also be off-target and we may loose the big picture (too strict). | |
| - review of ISO 19157 standard (version 2013) | |
| - See Table 1 for issues related to application of ISO quality elements to VGI | |
| - multiple aspects analyzed: | |
| - Data-based Indicators | |
| - Demographic and Socio-economic Indicators | |
| - Contributor Indicators | |
| - issues and challenges: | |
| - heterogeneity of the data and contributors | |
| - spatial bias | |
| - lack of specifications | |
| - dynamic nature in which the data are updated | |
| - patchiness of the contributions | |
| - lack of authoritative data | |
| ## Development and Completeness of Points Of Interest in Free and Proprietary Data Sets: A Florida Case Study | |
| ### bibtex | |
| @proceedings{ | |
| 0x002e6e4d, | |
| GOid = {0xc1aa5572_0x003e2ed3}, | |
| editor = {Thomas Jekel}, | |
| editor = {Adrijana Car}, | |
| editor = {Josef Strobl}, | |
| editor = {Gerald Griesebner (Eds.)}, | |
| subject = {geography}, | |
| language = {en}, | |
| isbn = {978-3-87907-532-4}, | |
| publisher = {Verlag der Österreichischen Akademie der Wissenschaften}, | |
| copyright = {Österreichische Akademie der Wissenschaften}, | |
| year = {2013}, | |
| pages = {39-48}, | |
| title = {Development and Completeness of Points Of Interest in Free and Proprietary Data Sets: A Florida Case Study}, | |
| author = {Hartwig Hochmair}, | |
| author = {Dennis Zielstra}, | |
| volume = {Volume 1}, | |
| URL = {https://austriaca.at/?arp=0x002e6e4d}, | |
| note = {Online available: https://austriaca.at/?arp=0x002e6e4d - Last access:23.1.2025}, | |
| address = {Wien}, | |
| } | |
| ### Location | |
| - Florida, USA | |
| ### Method of position accuracy assessment | |
| - count of POIs in OSM and compared with other data sources | |
| ### Summary keywords | |
| - data sources to compare with OSM in 2012: | |
| - Geographic Names Information System (GNIS): US | |
| - TIGER/Line: US | |
| - TomTom MultiNet | |
| - ESRI Data & Maps and Streetmap North America | |
| - NAVTEQ NAVSTREETS | |
| - compared only the POIs count for 15 selected OSM POI classes | |
| ## Data Quality of Points of Interest in Selected Mapping and Social Media Platforms | |
| ### bibtex | |
| @inbook{inbook, | |
| author = {Hochmair, Hartwig and Juhász, Levente and Cvetojevic, Sreten}, | |
| year = {2018}, | |
| month = {01}, | |
| pages = {293-313}, | |
| title = {Data Quality of Points of Interest in Selected Mapping and Social Media Platforms}, | |
| isbn = {978-3-319-71469-1}, | |
| doi = {10.1007/978-3-319-71470-7_15} | |
| } | |
| ### Location | |
| - 7 cities: | |
| - Albuquerque, NM, USA | |
| - Cairns, Australia | |
| - Gainseville, FL, USA | |
| - London, UK | |
| - Nairobi, Kenya | |
| - Qingdao, China | |
| - Salzburg, Austria (analyzed in detail) | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - sources: OSM, Google, Foursquare, Facebook, Instagram, Twitter, Yelp | |
| - used tweets from sept-oct 2016 using bounding boxes | |
| - problems with location spoofing or personal info (my bed, hell, boring school, etc.) with possibel religious or political messages manually reviewed when the amount of POI stacking was 15 or more (arbitrary number). | |
| - issues with language, especially with Google/yelp and others which returns different sets of POIs for the same query in different languages | |
| - analysis of the frequencies of pairs of tag-value in OSM in different areas | |
| - analysis of density of POIs in each dataset for each region using NNi analysis (nearest neighbor index) | |
| - simple analysis of poi accuracy (like only in/out of water ?!) with small sample verified in the Foursquare database (by super-user, not by authors) | |
| - data quality only analysed for Salzburg due to local knowledge by authors, some points selected to be verified manually by authors and corrected for most datasets | |
| - offsets differs by datasets, OSM and Google being the best with very small offsets while facebook, foursquare and instagram being the ones with the largest occurence of large offsets: | |
| ## A weighted multi-attribute method for matching user-generated Points of Interest | |
| ### bibtex | |
| @article{article, | |
| author = {McKenzie, Grant and Janowicz, Krzystof and Adams, Benjamin}, | |
| year = {2014}, | |
| month = {01}, | |
| pages = {125-137}, | |
| title = {A weighted multi-attribute method for matching user-generated Points of Interest}, | |
| volume = {41}, | |
| journal = {Cartography and Geographic Information Science}, | |
| doi = {10.1080/15230406.2014.880327} | |
| } | |
| ### Location | |
| - sample of 200 Yelp POIs in the continental USA which were manually matched with 140 Foursquare POIs | |
| - test set of 73304 POIs | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - conflation/merging of POIs datasets (two steps): | |
| - matching | |
| - attributes conflation/merging | |
| - focuses on matching Foursquare with Yelp (conflation mostly) | |
| - mean distance between matched POIs was 63m (max 870m) with manually matched POIs | |
| - used max 1000 radius for matching with larger datasets | |
| - used Levenstein distance and phonetic similarity for matching (two methods) | |
| - used distance as a third method | |
| - used LDA (Latent Dirichlet allocation) as the fourth method | |
| - most distances are bewteen 0 and 250m with some outliers with larger distances | |
| - using four methods to match results in significant improvements over using a single method in matching accuracy: up to 95% (models tested with 140 POIs) | |
| - using a regression-based weighted multi-attribute model to combine the four methods results in over 97% matching | |
| ## A travel demand modeling framework based on OpenStreetMap | |
| ### bibtex | |
| @article{article, | |
| author = {Notelaers, Lotte and Verstraete, Jeroen and Vansteenwegen, Pieter and Tampère, Chris}, | |
| year = {2024}, | |
| month = {06}, | |
| pages = {}, | |
| title = {A travel demand modeling framework based on OpenStreetMap}, | |
| volume = {1}, | |
| journal = {Discover Civil Engineering}, | |
| doi = {10.1007/s44290-024-00020-y} | |
| } | |
| ### Location | |
| - Antwerp, Belgium | |
| - Leuven, Belgium | |
| - Ghent, Belgium | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - tradional: using TAZ (traffic analysis zones), but not precise enough (aggregated data), we need more precision for micro-mobility and capture the complextities of travel patterns: using POIs instead of TAZ | |
| - proposes a methodology to use OSM data for travel demand modeling, including extracting, cleaning and processing OSM data | |
| - issues with attributes in OSM (a lot of tags, a lot of them unused, deprecated, or unclear, especially buildings with no additinal tags than building=yes) | |
| - produced a fully-automated python toolkit (Poidpy) for OSM data gathering, cleaning and categorization | |
| - approximations: | |
| - ignore buildings less than 40m2 | |
| - artificially increase the area of downtown buildings to account for undeclared building levels | |
| - ignore nodes and polygons with unrelated tags (not required for modeling) | |
| - infer building purpose based on land use (building=yes) | |
| - do not consider buildings > 600m2 as residential | |
| - consider buildings inside city center to be mixed-use | |
| - ... | |
| - comparison with ground-truth, which is considered to be the OD matrices from Flanders person model provided by the department of “Mobiliteit en Openbare Werken” (MOW) of the Flemish government | |
| - goal: predict the production and attraction of trips using the area and type of buildings/POIs as weights | |
| - results: strong correlation for production and attraction, but in the three cases: underestimation of trips in downtown | |
| - python toolkit source code available | |
| ## Graph-Based Matching of Points-of-Interest from Collaborative Geo-Datasets | |
| ### bibtex | |
| @Article{ijgi7030117, | |
| AUTHOR = {Novack, Tessio and Peters, Robin and Zipf, Alexander}, | |
| TITLE = {Graph-Based Matching of Points-of-Interest from Collaborative Geo-Datasets}, | |
| JOURNAL = {ISPRS International Journal of Geo-Information}, | |
| VOLUME = {7}, | |
| YEAR = {2018}, | |
| NUMBER = {3}, | |
| ARTICLE-NUMBER = {117}, | |
| URL = {https://www.mdpi.com/2220-9964/7/3/117}, | |
| ISSN = {2220-9964}, | |
| DOI = {10.3390/ijgi7030117} | |
| } | |
| ### Location | |
| - London, UK using bounding box (8238 POIs from OSM and 13548 from Foursquare) | |
| - 200 POIs from OSm used aas test dataset | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - lack of gold standard to know what is the ground-truth | |
| - proposes a mthod to match two VGI databases without reliaing on training data | |
| - usual stesp for matching: | |
| - pre-proce | |
| - candidate selection | |
| - computation of similarity measures | |
| - aggregation of similarity measures | |
| - matching decision | |
| - matching evaluation | |
| - three different similarity measures: | |
| - spatial (euclidian distance), string/name (Token Sort Ratio/Toekn Set Ratio) and semantic similarities (WordNet in English + Lin Mesure) | |
| - see Lin, D. An information-theoretic definition of similarity. In Proceedings of the 15th International Conference on Machine Learning, Madison, WI, USA, 24–27 July 1998. | |
| - graph-based matching: | |
| - nodes: POIs | |
| - edges: possibility that two or more POIs are the same with weight | |
| - distance differences mostly < 50m with some outliers up to 90m | |
| - accuracy of 86% with test data with best method | |
| - contribution: fast, simple, effective and transferable | |
| ## The quality of OpenStreetMap food-related point-of-interest data for use in epidemiological research | |
| ### bibtex | |
| @article{article, | |
| author = {Matias de Pinho, Maria Gabriela and Flueckiger, Benjamin and Valentin, Antonia and Kasdagli, Maria-Iosifina and Kyriakou, Kalliopi and Lakerveld, Jeroen and Mackenbach, Joreintje and Beulens, Joline and Hoogh, Kees de}, | |
| year = {2023}, | |
| month = {07}, | |
| pages = {103075}, | |
| title = {The quality of OpenStreetMap food-related point-of-interest data for use in epidemiological research}, | |
| volume = {83}, | |
| journal = {Health & place}, | |
| doi = {10.1016/j.healthplace.2023.103075} | |
| } | |
| ### Location | |
| - five European regions (June 2020): | |
| - Switzerland | |
| - Greece | |
| - Spain | |
| - The Netherlands | |
| - Poland | |
| ### Method of position accuracy assessment | |
| - Google Street-View images | |
| ### Summary keywords | |
| - specialized to food-related OSM data | |
| - used Google street view to verify data (5% of all eligible street segments in cities from the 5 regions) | |
| - local data used as reference | |
| - analyzed accessibility to food-related POIs with distinct buffers by type of POI | |
| - count of POIs in OSM significantly lower than in reference (incomplete) | |
| - could only compare for three of the 5 regions, because only 3 had a trustworthy reference dataset | |
| ## There's No Such Thing as the Perfect Map - Quantifying Bias in Spatial Crowd-sourcing Datasets | |
| ### bibtex | |
| @inproceedings{inproceedings, | |
| author = {Quattrone, Giovanni and Capra, Licia and De Meo, Pasquale}, | |
| year = {2015}, | |
| month = {02}, | |
| pages = {1021-1032}, | |
| title = {There's No Such Thing as the Perfect Map}, | |
| doi = {10.1145/2675133.2675235} | |
| } | |
| ### Location | |
| - 40 countries between 2010 and 2012 (POIs only) | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - important issue: content bias | |
| - measure the misalignment between OSM contributors and power users (organizations using OSM for their own location-based services) | |
| - results: | |
| - very low bias in terms of waht content is being produced | |
| - higher bias in terms of where information is being mapped | |
| - bias on how meticulously the map is being edited to vary with culture: where power distance is low and individualism is high, the crowd performs a more accurate work than power users, and viceversa. | |
| ## Crowdsourcing, Citizen Science or Volunteered Geographic Information? The Current State of Crowdsourced Geographic Information | |
| ### bibtex | |
| @article{article, | |
| author = {See, Linda and Mooney, Peter and Foody, Giles and Bastin, Lucy and Comber, Alexis and Estima, Jacinto and Fritz, Steffen and Kerle, Norman and Jiang, Bin and Laakso, Mari and Liu, Hai-Ying and Milcinski, Grega and Niksic, Matej and Pődör, Andrea and Olteanu Raimond, Ana-Maria and Rutzinger, Martin}, | |
| year = {2016}, | |
| month = {04}, | |
| pages = {}, | |
| title = {Crowdsourcing, Citizen Science or Volunteered Geographic Information? The Current State of Crowdsourced Geographic Information}, | |
| volume = {5}, | |
| journal = {ISPRS International Journal of Geo-Information}, | |
| doi = {10.3390/ijgi5050055} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - first objective: present a compilation of terms, providing some basic definitions and their primary attributions about VGI (review of terminology): see Figure 1 | |
| - automated literature review of 25338 papers based on keywords (terminology) | |
| - Google trends analysis of search terms (terminology) | |
| - second objective: better understand the current state of mapping and spatial data collection by citizens through a systematic review of different online initiatives | |
| - used VGI-Net | |
| - analyzed quality and use of data for research | |
| - quality control is hard to evaluate | |
| - analysis of participants | |
| ## Assessing OpenStreetMap Data Using Intrinsic Quality Indicators: An Extension to the QGIS Processing Toolbox | |
| ### bibtex | |
| @Article{fi9020015, | |
| AUTHOR = {Sehra, Sukhjit Singh and Singh, Jaiteg and Rai, Hardeep Singh}, | |
| TITLE = {Assessing OpenStreetMap Data Using Intrinsic Quality Indicators: An Extension to the QGIS Processing Toolbox}, | |
| JOURNAL = {Future Internet}, | |
| VOLUME = {9}, | |
| YEAR = {2017}, | |
| NUMBER = {2}, | |
| ARTICLE-NUMBER = {15}, | |
| URL = {https://www.mdpi.com/1999-5903/9/2/15}, | |
| ISSN = {1999-5903}, | |
| DOI = {10.3390/fi9020015} | |
| } | |
| ### Location | |
| - Case study: Punjab, India OSM Data (2016-2017) | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - objective: extand QGIS toolbox for: | |
| - network length completeness | |
| - attribute completeness | |
| - semantic accuracy | |
| - heuristic iindicator to assess route navigability | |
| - quality of data in VGI/OSM: a real challenge | |
| ## A review of volunteered geographic information quality assessment methods | |
| ### bibtex | |
| @article{article, | |
| author = {Senaratne, Hansi and Mobasheri, Amin and Ali, Ahmed and Cristina, Capineri and Haklay, Muki}, | |
| year = {2016}, | |
| month = {05}, | |
| pages = {}, | |
| title = {A review of volunteered geographic information quality assessment methods}, | |
| volume = {31}, | |
| journal = {International Journal of Geographical Information Science}, | |
| doi = {10.1080/13658816.2016.1189556} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - data mining as one more stand alone approach to assess VGI quality (discover patterns) | |
| - quality measures (see ISO standards) and quality indicators (when authoritative data is unavailable) | |
| - map-based (like OSM), image-based VGI (like Flickr) and text-based VGI (like Twitter, Reddit, etc.) | |
| - literature review of papers about quality in VGI based on these attributes: | |
| - Positional accuracy | |
| - Thematic accuracy | |
| - Topological consistency | |
| - Completeness | |
| - Temporal accuracy | |
| - Geometric accuracy | |
| - Semantic accuracy | |
| - Lineage | |
| - Usage | |
| - Credibility | |
| - Trustworthiness | |
| - Content quality | |
| - Vagueness | |
| - Local knowledge | |
| - Experience | |
| - Recognition | |
| - Reputation | |
| - types of approcahes and methods: | |
| - Geographic | |
| - Social | |
| - Crowdsourcing | |
| - Data mining | |
| ## Using OpenStreetMap point-of-interest data to model urban change—A feasibility study | |
| ### bibtex | |
| @article{10.1371/journal.pone.0212606, | |
| doi = {10.1371/journal.pone.0212606}, | |
| author = {Zhang, Liming AND Pfoser, Dieter}, | |
| journal = {PLOS ONE}, | |
| publisher = {Public Library of Science}, | |
| title = {Using OpenStreetMap point-of-interest data to model urban change—A feasibility study}, | |
| year = {2019}, | |
| month = {02}, | |
| volume = {14}, | |
| url = {https://doi.org/10.1371/journal.pone.0212606}, | |
| pages = {1-34}, | |
| number = {2}, | |
| } | |
| ### Location | |
| - foursquare coffee shop data used as reference (851 POIs, from June 2018) | |
| - osm coffee shop data matching foursquare data (529 matching POIs) | |
| - Manhattan, New York | |
| ### Method of position accuracy assessment | |
| - compare between sources (osm and foursquare), no ground-truth reference | |
| - Levenshtein distance for text matching | |
| ### Summary keywords | |
| - objective: assess the suitability of user-generated content in the form of OpenStreetMap POI data as a means to infer urban change | |
| - accuracy and coverage | |
| - trend worthiness (assess whether VGI can match some commonly recognized trends or theories): compare coffee shop POI data with urban housing pricing | |
| - compare OSM with Foursquare | |
| - built models incremently and compared with Akaike Information Crite- rion (AIC) and Bayesian Information Criterion (BIC) | |
| - only 35% can be matched exactly within 50m, but 60% can be matched using spatial and text matching | |
| ## A Guide to Geospatial Data Quality | |
| ### bibtex | |
| @presentation{presentation, | |
| author = {Bédard, Yvan and Wright, Eric and Rivest, Sonia}, | |
| year = {2015}, | |
| month = {06}, | |
| pages = {}, | |
| title = {A Guide to Geospatial Data Quality}, | |
| doi = {10.13140/RG.2.1.2819.1763} | |
| } | |
| ### Location | |
| ### Method of position accuracy assessment | |
| ### Summary keywords | |
| - See below, main section: ISO 19157 (2023) | |
| ## DO NOT FORGET TO TALK ABOUT THIS: | |
| - Apple Maps (not analyzed/compared): only available for Apple devices and some windows browsers | |
| - Google Maps leads significantly with over a billion monthly active users. Apple Maps has made notable progress, claiming between 500 million monthly users. Bing sees around 100 million daily visitors, with app usage increased by six times and with a 6 time increase (including Bing Maps), and Bing Maps being used by over 19,000 live websites. (Find source) | |
| - Alignement problems: that's why we use a reference geodesic aligned aerial imagery dataset from Quebec ministry of natural resources | |
| - What is considered a large enough distance between truth and provided data so that it would affect travel behaviour and simulation results? Especially for slow and active modes like walking and cycling, but also transit accessibility, including the choice of the nearest or most efficient bus stop locations. | |
| - Data on where the POIs are approximately is widely available, but the precise location is not. We don't really know where the POIs are exactly. Most of the time, reference datasets are maped using centroids of buildings and/or the whole land, which may not be precise enough to analyse and simulate correctly travel times and accessibility to bus stops, bicycle infrastructre or pedestrian networks, especially for rural areas or large factories/parks/shopping malls/university campuses/large office buildings, etc. | |
| - Even if differences in distances between validated POI entrance location and google/osm/landrole/overture data sources are small most of the time, since there is a tendency to map POI in the centroid of their buildings (and landrole in the centroid of the whole land), these differences in distances translate most of the time a net increase in walking/cycling travel time, which is not always negligeable, especially for elderly people, people with disabilities or young children with their parents, and thus these distances may not exist in reality since entrances can be closer to the main road then the building or land centroid. In that sense, accessibiility may be even better in reality than what is calculated with the data sources usually used. | |
| - Use a graph like in Yeow 2021 - Point-of-Interest POI Data Validation Methods An Urban Case Study: figure 3 for distance comparisons | |
| - Explain the ISO 19157 elements (see main section below: ISO 19157 (2023)) | |
| # ISO 19157 (2023) | |
| From Bédard 2015 - A Guide to Geospatial Data Quality - ISO 19157 untile we can get access to the ISO standard. | |
| ISO 19157 Geographic information – Data quality: establishes | |
| principles for describing the quality of geographic data by: | |
| - Defining components for describing data quality | |
| - Specifying components and content structure of a register for data quality | |
| measures | |
| - Describing general procedures for evaluating the quality of geographic | |
| data | |
| - Establishing principles for reporting data quality | |
| ## Internal quality | |
| - Completeness: presence or absence of features, their attributes | |
| - Omission or commission | |
| - Logical consistency: degree of adherence to logical rules of data | |
| structure, attribution and relationships | |
| - Conceptual consistency, domain consistency, format consistency, | |
| - Positional accuracy: accuracy of the position of features within a | |
| spatial reference system | |
| - Absolute accuracy, relative accuracy, gridded data position accuracy | |
| - Thematic accuracy: quality of the thematic attributes and of the | |
| classifications of features and their relationships | |
| - Classification correctness, non quantitative attribute correctness, | |
| quantitative attribute accuracy | |
| - Temporal quality: quality of the temporal attributes and temporal | |
| relationships of features | |
| - Accuracy of time measurement, temporal consistency, temporal | |
| validity | |
| ## External quality | |
| - Fitness for use: degree of agreement between data | |
| characteristics (i.e., internal quality) and the explicit and/or | |
| implicit needs of a user for a given application in a given | |
| context | |
| - Usability: based on user’s requirements, all internal quality | |
| elements of ISO 19157 may be used to evaluate usability | |
| ## Perceived quality | |
| - Within a consumer-centered (i.e. B2C and C2C), users may | |
| have a different view of the external data quality of a | |
| geospatial dataset | |
| - Users can rate the dataset based on their perception (using | |
| a 5-star system for example) and write comments | |
| - The global perceived quality is the result of the aggregation | |
| of each individual user perception (bottom-up approach) | |
| ## Metaquality (ISO 19157) | |
| - Information describing the quality of data quality | |
| information: | |
| - Confidence | |
| - Representativity | |
| - Homogeneity | |
| - Metaquality helps estimating the risk related to geospatial | |
| data uncertainties | |
| ## Data quality scopes | |
| Geospatial data quality can be regarded at various granularity levels | |
| (ISO 19157 data quality scopes) | |
| - Dataset series level (e.g., the National Topographic System of Canada | |
| (NTS)) | |
| - Dataset level (e.g., a specific map of the NTS) | |
| - Subset level (e.g., the subset of features included in the North-West | |
| zone) | |
| - Feature type level (e.g., the set of “roads segments” of a topographic | |
| map) | |
| - Feature instance level (e.g., road 138 on a specific topographic map) | |
| - Feature attribute level (e.g., the “Functional road class” of a road | |
| segment) | |
| - Attribute value level (e.g., code value “1” (“Freeway”) for a specific | |
| road segment) | |
| ISO 19157 provides methods for aggregating quality information from | |
| a single data of a single feature up to the complete dataset | |
| ## Risk | |
| Risk is about the effect of uncertainty: | |
| - Uncertain geospatial data | |
| - Uncertain geospatial data quality | |
| - Uncertain geospatial data usages | |
| - Uncertain expertise of users of geospatial data | |
| # Web references to read and/or include in the paper: | |
| ## OSM and crowd sourced mapping data | |
| - [Volunteered Geographic Information](https://en.wikipedia.org/wiki/Volunteered_geographic_information) | |
| - [OpenStreetMap](https://www.openstreetmap.org/) | |
| - [Point of Interest](https://en.wikipedia.org/wiki/Point_of_interest) | |
| - [Areas of Interest OSM github](https://github.com/geometalab/aoi-osm) | |
| - [Geofabrik Map Compare](https://tools.geofabrik.de/mc/) | |
| ## Google Maps | |
| - [Google Maps (Wikipedia)](https://en.wikipedia.org/wiki/Google_Maps) | |
| - [Google Maps](https://www.google.com/maps) | |
| - [Google Maps API](https://developers.google.com/maps) | |
| - [Geocoding API](https://en.wikipedia.org/wiki/Geocoding_API) | |
| - [Geocoding Service](https://en.wikipedia.org/wiki/Geocoding_service) | |
| - [Google Maps Places API](https://developers.google.com/maps/documentation/places/web-service/overview) | |
| - [Google Maps 101: How We Map the World](https://www.blog.google/products/maps/google-maps-101-how-we-map-world/) | |
| ## Overture | |
| - [Overture Maps](https://overturemaps.org/) | |
| - [Overture Maps API](https://docs.overturemaps.org/guides/places/) | |
| - [Overture Maps Foundation](https://en.wikipedia.org/wiki/Overture_Maps_Foundation) | |
| - [Overture Maps Explore](https://explore.overturemaps.org/#15/38.90678/-77.03649) | |
| - [Overture Maps Data](https://overturemaps.org/data/) | |
| ## Other | |
| - [Geocoding](https://en.wikipedia.org/wiki/Address_geocoding) | |
| - [The Battle of the Maps](https://www.techwyse.com/blog/digital-marketing-101/the-battle-of-the-maps-google-vs-bing-vs-apple-in-2024#) | |
| - [Navigation App Market](https://www.businessofapps.com/data/navigation-app-market/) | |
| - [Bing Statistics](https://coalitiontechnologies.com/blog/bing-statistics-search-and-usage-data-in-2024) | |
| - [The Battle of the Maps](https://www.techwyse.com/blog/digital-marketing-101/the-battle-of-the-maps-google-vs-bing-vs-apple-in-2024#) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment