From 5fd7d4bc63f17586986ca26ab12ab1730615ea2d Mon Sep 17 00:00:00 2001
From: Franck Dary <franck.dary@lis-lab.fr>
Date: Wed, 23 Oct 2019 12:16:55 +0200
Subject: [PATCH] Improved lemmatization
---
transition_machine/src/Oracle.cpp | 15 +++++++++++++++
1 file changed, 15 insertions(+)
diff --git a/transition_machine/src/Oracle.cpp b/transition_machine/src/Oracle.cpp
index 097a7a3..a603532 100644
--- a/transition_machine/src/Oracle.cpp
+++ b/transition_machine/src/Oracle.cpp
@@ -406,6 +406,11 @@ void Oracle::createDatabase()
newState = "lemmatizer_rules";
movement = 0;
}
+ else if (previousAction == "nothing")
+ {
+ newState = "lemmatizer_lookup";
+ movement = 1;
+ }
else
{
newState = "lemmatizer_lookup";
@@ -676,6 +681,16 @@ void Oracle::createDatabase()
const std::string & form = c.getTape("FORM")[0];
const std::string & pos = c.getTape("POS")[0];
std::string lemma;
+
+ if (c.hasTape("ID"))
+ {
+ auto & id = c.getTape("ID")[0];
+ if (!id.empty())
+ {
+ if (util::split(id, '-').size() > 1)
+ return std::string("NOTHING");
+ }
+ }
if (oracle->data.count(form + "_" + pos))
lemma = oracle->data[form + "_" + pos];
--
GitLab