Hulan-2014-10-31 - 版本历史

Lr：/* improve lucene search */

2014-10-31T05:32:33Z

‎improve lucene search

Lr：/* Algorithm */

2014-10-31T05:05:14Z

‎Algorithm

2014年10月31日 (五) 05:04 Lr

2014-10-31T05:04:45Z

Lr：/* plan to discuss */

2014-10-31T05:04:13Z

‎plan to discuss

Lr：/* Knowledge Management and labeling system */

2014-10-31T05:03:19Z

‎Knowledge Management and labeling system

2014年10月31日 (五) 05:02 Lr

2014-10-31T05:02:40Z

2014年10月31日 (五) 04:59 Lr

2014-10-31T04:59:48Z

Lr：/* improve lucene search */

2014-10-31T04:53:43Z

‎improve lucene search

Lr：以“==Dialog system== ==Algorithm== ===Spell mistake=== :* using ngram to get candidate sentence. ===improve lucene search=== * lucene similarity method {| border="2px"...”为内容创建页面

2014-10-31T04:50:17Z

以“==Dialog system== ==Algorithm== ===Spell mistake=== :* using ngram to get candidate sentence. ===improve lucene search=== * lucene similarity method {| border="2px"...”为内容创建页面

新页面

==Dialog system==
==Algorithm==
===Spell mistake===
:* using ngram to get candidate sentence.
===improve lucene search===
* lucene similarity method
{| border="2px"
|+ different result in lucene
|-
! method !!Default !! BM25 !! LMDirichlet !! DFR !! LMJelinekMercer !! IB
|-
! Accary
| 0.66228 || 0.66228 || 0.4091 || 0.65476 || 0.65476 || 0.6666
|-
|}

* our vsm method
:* our vsm method re-rank(54%),lucene(67%)

* lucene top50(caoli)
:* top10(82.95%),top20(86.34),top50(90.22%)
:* need to check the other 10% error

* lucene Optimization(liurong)
:* rewrite the method to select the 50 standard question not same template.
:* test the boost keyword weight and extract the synonyms word.
:* check the word segment for template.
:* min-segment method improve the accuracy.

==plan to discuss==

@@ 第2行： / 第2行： @@
 ==Algorithm==
 ===Spell mistake===
-:* using ngram to get candidate sentence.
+:* using ngram to get candidate sentence.(xingchao)
 ===improve lucene search===
 * lucene similarity method
@@ 第32行： / 第32行： @@
 :* test the different idf vale from baidu sougou in fuzzymatch.
 :* IDF from train-data performance bad than default IDF,from 0.63->0.69.
 ==knowledge structure==
 * structure the default answer using attributes of the entity.

←上一版本		2014年10月31日 (五) 05:04的版本
第38行：		第38行：

	==plan to discuss==		==plan to discuss==
		+	* add the triples search to QA engine

@@ 第35行： / 第35行： @@
 * structure the default answer using attributes of the entity.
 ==Knowledge Management and labeling system==
-:* prepare the interface and function.
+* prepare the interface and function.
 ==plan to discuss==

@@ 第26行： / 第26行： @@
 :* test the boost keyword weight and extract the synonyms word.
 :* check the word segment for template.
-:* min-segment method improve the accuracy.
+:* min-segment method improve the accuracy.(0.61->0.66)
 :* check the query method for getting lucene information and to rewrite the score method like the idf value.
@@ 第34行： / 第34行： @@
 ==knowledge structure==
 * structure the default answer using attributes of the entity.
 ==plan to discuss==

@@ 第29行： / 第29行： @@
 :* check the query method for getting lucene information and to rewrite the score method like the idf value.
-*IDF
+*IDF(caoli)
 :* test the different idf vale from baidu sougou in fuzzymatch.
+:* IDF from train-data performance bad than default IDF,from 0.63->0.69.
 ==plan to discuss==

@@ 第16行： / 第16行： @@
 * our vsm method
-:* our vsm method re-rank(54%),lucene(67%)
+:* our vsm method re-rank(54%),lucene(66.28%)
 * lucene top50(caoli)

@@ 第1行： / 第1行： @@
-==Dialog system==
+=Dialog system=
 ==Algorithm==
 ===Spell mistake===