@article{fdi:010063972,
  title = {{S}tructural knowledge learning from maps for supervised land cover/use classification : application to the monitoring of land cover/use maps in {F}rench {G}uiana},
  author = {{B}ayoudh, {M}. and {R}oux, {E}mmanuel and {R}ichard, {G}. and {N}ock, {R}.},
  editor = {},
  language = {{ENG}},
  abstract = {{T}he number of satellites and sensors devoted to {E}arth observation has become increasingly elevated, delivering extensive data, especially images. {A}t the same time, the access to such data and the tools needed to process them has considerably improved. {I}n the presence of such data flow, we need automatic image interpretation methods, especially when it comes to the monitoring and prediction of environmental and societal changes in highly dynamic socio-environmental contexts. {T}his could be accomplished via artificial intelligence. {T}he concept described here relies on the induction of classification rules that explicitly take into account structural knowledge, using {A}leph, an {I}nductive {L}ogic {P}rogramming ({ILP}) system, combined with a multi-class classification procedure. {T}his methodology was used to monitor changes in land cover/use of the {F}rench {G}uiana coastline. {O}ne hundred and fifty-eight classification rules were induced from 3 diachronic land cover/use maps including 38 classes. {T}hese rules were expressed in first order logic language, which makes them easily understandable by non-experts. {A} 10-fold cross-validation gave significant average values of 84.62%, 99.57% and 77.22% for classification accuracy, specificity and sensitivity, respectively. {O}ur methodology could be beneficial to automatically classify new objects and to facilitate object-based classification procedures.},
  keywords = {{S}upervised classification ; {M}achine learning ; {I}nductive {L}ogic {P}rogramming ({ILP}) ; {G}eographic {I}nformation {S}ystem ; {L}and cover map ; {GUYANE} {FRANCAISE}},
  booktitle = {},
  journal = {{C}omputers and {G}eosciences},
  volume = {76},
  numero = {},
  pages = {31--40},
  ISSN = {0098-3004},
  year = {2015},
  DOI = {10.1016/j.cageo.2014.08.013},
  URL = {https://www.documentation.ird.fr/hor/fdi:010063972},
}