我正在使用oracle text for Arabic language。
我想将同义词列表保存在数据库表中,所以从这个表中读取域索引,有什么想法吗?
答案 0 :(得分:1)
我找到了解决方案:
1-我将我的同义词列表上传到一个名为words的表(包含所有术语及其同义词' ID)和名为synset的主表(包含同义词)
2-创建词库:
begin
ctx_thes.create_thesaurus ('MyThesaurus');
end;
3-创建一个存储过程以从我的表[words]中读取并创建同义词之间的关系:
create or replace procedure CreateSynonyms is
CURSOR syn_cur is select s.name_abstract,w.root,w.word_abstract
from p words w , synset s
where w.synset_id=s.synset_id and w.root<>s.name_abstract and w.word_abstract<> s.name_abstract
order by s.synset_id;
syn_rec syn_cur%rowtype;
BEGIN
OPEN syn_cur;
LOOP
FETCH syn_cur into syn_rec;
EXIT WHEN syn_cur%notfound;
begin
ctx_thes.create_relation ('MyThesurus', syn_rec.name_abstract, 'syn', syn_rec.word_abstract);
END LOOP;
END;
4-重写我的查询以选择同义词:
select /*+ FIRST_ROWS(1)*/ sentence_id,score(1) as sc, isn
where contains(PROCESSED_TEXT,'<query>
<textquery>
search for somthing here
<progression>
<seq><rewrite>transform((TOKENS, "{", "}", ","))</rewrite></seq>
<seq><rewrite>transform((TOKENS, "syn(", ",listing)", " , "))</rewrite>/seq>
</progression>
</textquery>
<score datatype="INTEGER" algorithm="COUNT"/></query>',1)>0
希望这会有所帮助