fr.univnantes.termsuite.resources.default-extractor-config.json Maven / Gradle / Ivy

Go to download

Show more of this group Show more artifacts with this name
Show all versions of termsuite-core Show documentation

A Java UIMA-based toolbox for multilingual and efficient terminology extraction an multilingual term alignment

The newest version!

{
  "pre-filter" : {
    "enabled" : false,
    "type" : "THRESHOLD",
    "property" : "FREQUENCY",
    "top-n" : 500,
    "threshold" : 2.0,
    "keep-variants" : false,
    "max-variant-num" : 25
  },
  "contextualizer" : {
    "enabled" : false,
    "scope" : 3,
    "coocc-frequency-th" : 2,
    "association-rate" : "fr.univnantes.termsuite.engines.contextualizer.MutualInformation"
  },
  "morphology" : {
    "enabled" : true,
    "alpha" : 0.7,
    "beta" : 0.1,
    "gamma" : 0.1,
    "delta" : 0.1,
    "prefix-splitter-enabled" : true,
    "derivatives-detector-enabled" : true,
    "native-splitter-enabled" : true,
    "min-component-size" : 3,
    "max-component-num" : 3,
    "score-th" : 0.85,
    "segment-similarity-th" : 1.0
  },
  "gatherer" : {
    "semantic-enabled" : false,
    "semantic-similarity-distance" : "fr.univnantes.termsuite.metrics.Cosine",
    "semantic-similarity-th" : 0.3,
    "semantic-nb-candidates" : 5,
    "semantic-dico-only" : false,
    "graphical-enabled" : true,
    "graphical-similarity-th" : 0.95,
    "max-class-size" : 2000,
    "max-num-comparisons" : 1000000
  },
  "post-processor" : {
    "min-swt-length" : 4,
    "affix-score-th" : 0.25,
    "variation-score-th" : 0.2,
    "orthographic-score-th" : 0.55,
    "term-independance-th" : 0.1
  },
  "post-filter" : {
    "enabled" : false,
    "type" : "TOP_N",
    "property" : "SPECIFICITY",
    "top-n" : 500,
    "threshold" : 2.0,
    "keep-variants" : true,
    "max-variant-num" : 25
  },
  "ranking" : {
    "property" : "SPECIFICITY",
    "is-descending" : true
  }
}