Basic coref

app.py

@@ -16,6 +16,7 @@ st.title('Entity Linking Demo')
 article = st.text_area('Article to analyze:', value=open("example.txt").read())
 
 seen_entities = []
+seen_surnames = []
 if st.button('Submit'):
     good_ents = []
 
@@ -24,6 +25,16 @@ if st.button('Submit'):
         for ent in doc.ents:
             if ent._.kb_qid is None or ent.label_ not in ["ORG", "PERSON", "GPE"] or ent.text in seen_entities:
                 continue
+
+            if ent.label_ == "PERSON":
+                if len(ent.text.split()) == 1:
+                    # Single name
+                    if ent.text in seen_surnames:
+                        continue
+                else:
+                    # Multipart name
+                    seen_surnames.append(ent.text.split()[-1])
+
             seen_entities.append(ent.text)
             print((ent.text, ent.label_, ent._.kb_qid, ent._.url_wikidata, ent._.nerd_score))
             r = requests.get("https://www.wikidata.org/w/api.php?action=wbgetclaims&format=json&property=P18&entity=" + ent._.kb_qid)