marcverhagen / corelex

1 stars 0 forks source link

Review basic types #13

Open marcverhagen opened 2 years ago

marcverhagen commented 2 years ago

Basic types were adjusted for Wordnet 3.1, but I should check for each basic type whether new synsets should be added, either because new synsets were added to Wordnet or because the original set of basic types for Wordnet 1.5 missed some.

For example, 'chicken' now shows up with three basic types: sub, anm, hum and evt. One of the four senses has the following path to the toptype:

<Synset 07660576 n chicken.13.0 poulet.13.0 volaille.13.0 sub>
  <Synset 07660315 n poultry.13.0 sub>
    <Synset 07659991 n bird.13.0 fowl.13.0 sub>
      <Synset 07665463 n meat.13.1 sub>
        <Synset 07571428 n food.13.0 solid_food.13.0 sub>
          <Synset 15071467 n solid.27.4 sub>
            <Synset 00021007 n matter.03.0 *sub>
              <Synset 00001930 n physical_entity.03.0 *pho>
                <Synset 00001740 n entity.03.0 *ent>

So it ends up with sub as the basic type. Currently the fod basic type is defined as [('00011263', 'food nutrient')]. We should add ('07571428', 'food solid_food') andmaybe others as well.

For each basic type, do a Wordnet search of elements in the synsets used.