Merge cfe657d into aa061e5

alexandrainst · Jan 7, 2021 · f95a525 · f95a525
2 parents aa061e5 + cfe657d
commit f95a525
Show file tree

Hide file tree

Showing 3 changed files with 25 additions and 3 deletions.
diff --git a/danlp/datasets/dannet.py b/danlp/datasets/dannet.py
@@ -140,6 +140,24 @@ def hyponyms(self, word, pos=None):
 
         return hypernyms
 
+    def domains(self, word, pos=None):
+        """
+        Returns the domains of `word`.
+
+        :param word: text
+        :param pos: (list of) part of speech tag(s) (in "Noun", "Verb", "Adjective")
+        :return: list of domains
+
+        """
+
+        word_synset_ids = self._synset_ids(word, pos)
+        dom_synset_ids = self.relations[self.relations['synset_id'].isin(word_synset_ids) & (self.relations['relation']=='domain')]['value'].tolist()
+        dom_synset_ids = [val for val in dom_synset_ids if val.isdigit()]
+        domains_ids = self.wordsenses[self.wordsenses['synset_id'].isin(dom_synset_ids)]['word_id'].tolist()
+        domains = self.words[self.words['word_id'].isin(domains_ids)]['form'].tolist()
+
+        return domains
+
     def wordnet_relations(self, word, pos=None, eurowordnet=True):
         """
         Returns the name of the relations `word` is associated with.
@@ -185,14 +203,14 @@ def _synset_ids(self, word, pos=None):
 
     def _word_from_id(self, word_id):
 
-        assert(type(word_id) == int or (type(word_id) == str and word_id.is_digit()))
+        assert(type(word_id) == int or (type(word_id) == str and word_id.isdigit()))
         word_id = str(word_id)
 
         return self.words[self.words['word_id'] == word_id]['form'].tolist()
 
     def _synset_from_id(self, synset_id):
 
-        assert(type(synset_id) == int or (type(synset_id) == str and synset_id.is_digit()))
+        assert(type(synset_id) == int or (type(synset_id) == str and synset_id.isdigit()))
         synset_id = str(synset_id)
 
         synset_labels = self.synsets[self.synsets['synset_id'] == synset_id]['label'].tolist()

diff --git a/docs/docs/datasets.md b/docs/docs/datasets.md
@@ -176,7 +176,7 @@ dannet = DanNet()
 words, wordsenses, relations, synsets = dannet.load_with_pandas()
 ```
 
-We also provide helper functions to search for synonyms, hyperonyms and hyponyms through the databases. 
+We also provide helper functions to search for synonyms, hyperonyms, hyponyms and domains through the databases. 
 Once you have downloaded the DanNet wrapper, you can use the following features: 
 
 ```python
@@ -191,6 +191,9 @@ dannet.hypernyms(word)
 # hyponyms
 dannet.hyponyms(word)
 """ ['hærmyre', 'skovmyre', 'pissemyre', 'tissemyre'] """
+# domains
+dannet.domains(word)
+""" ['zoologi'] """
 # meanings
 dannet.meanings(word)
 """ ['ca. 1 cm langt, årevinget insekt med en kraftig in ... (Brug: "Myrer på terrassen, og andre steder udendørs, kan hurtigt blive meget generende")'] """

diff --git a/tests/test_datasets.py b/tests/test_datasets.py
@@ -176,6 +176,7 @@ def test_dannet(self):
         self.assertEqual(dannet.synonyms('kat'), ['missekat', 'mis'])
         self.assertEqual(dannet.hypernyms('myre', pos=['Noun']), ['årevingede insekter'])
         self.assertEqual(dannet.hyponyms('myre', pos='Noun'), ['hærmyre', 'skovmyre', 'pissemyre', 'tissemyre'])
+        self.assertEqual(dannet.domains('myre', pos='Noun'), ['zoologi'])
         self.assertEqual(dannet.pos('myre'), ['Noun'])
         self.assertEqual(dannet.meanings('myre'), ['ca. 1 cm langt, årevinget insekt med en kraftig in ... (Brug: "Myrer på terrassen, og andre steder udendørs, kan hurtigt blive meget generende")'])
         self.assertEqual(dannet.wordnet_relations('kat'), {'domain', 'has_holo_part', 'eq_has_synonym', 'has_hyperonym', 'role_agent', 'used_for', 'has_mero_part'})