Approaches for the clustering of geographic metadata and the automatic detection of quasi-spatial dataset series

Lacasta, J. ; López Pellicer, F. J. ; Zarazaga-Soria, J. ; Béjar, R. ; Nogueras-Iso, J.

Resumen: The discrete representation of resources in geospatial catalogues affects their information retrieval performance. The performance could be improved by using automatically generated clusters of related resources, which we name quasi-spatial dataset series. This work evaluates whether a clustering process can create quasi-spatial dataset series using only textual information from metadata elements. We assess the combination of different kinds of text cleaning approaches, word and sentence-embeddings representations (Word2Vec, GloVe, FastText, ELMo, Sentence BERT, and Universal Sentence Encoder), and clustering techniques (K-Means, DBSCAN, OPTICS, and agglomerative clustering) for the task. The results demonstrate that combining word-embeddings representations with an agglomerative-based clustering creates better quasi-spatial dataset series than the other approaches. In addition, we have found that the ELMo representation with agglomerative clustering produces good results without any preprocessing step for text cleaning.

Idioma: Inglés

DOI: 10.3390/ijgi11020087

Año: 2022

Publicado en: ISPRS International Journal of Geo-Information 11, 2 (2022), 87[19 pp.]
ISSN: 2220-9964

Factor impacto JCR: 3.4 (2022)

Categ. JCR: GEOGRAPHY, PHYSICAL rank: 18 / 49 = 0.367 (2022) - Q2 - T2
Categ. JCR: COMPUTER SCIENCE, INFORMATION SYSTEMS rank: 88 / 158 = 0.557 (2022) - Q3 - T2
Categ. JCR: REMOTE SENSING rank: 21 / 34 = 0.618 (2022) - Q3 - T2

Factor impacto CITESCORE: 6.2 - Earth and Planetary Sciences (Q1) - Social Sciences (Q1)

Factor impacto SCIMAGO: 0.738 - Earth and Planetary Sciences (miscellaneous) (Q1) - Geography, Planning and Development (Q1) - Computers in Earth Sciences (Q2)

Financiación: info:eu-repo/grantAgreement/ES/AEI/PID2020-113353RB-I00
Financiación: info:eu-repo/grantAgreement/ES/DGA/T59-20R

Tipo y forma: Artículo (Versión definitiva)