2014-11-25 - 版本历史

Lr：/* Text Processing */

2014-12-08T02:00:36Z

‎Text Processing

Lr：/* Text Processing */

2014-12-01T07:01:32Z

‎Text Processing

2014年11月24日 (一) 11:43 Zhangzy

2014-11-24T11:43:03Z

2014年11月24日 (一) 11:42 Zhangzy

2014-11-24T11:42:36Z

2014年11月24日 (一) 11:39 Zhangzy

2014-11-24T11:39:49Z

2014年11月24日 (一) 08:54 Yinshi

2014-11-24T08:54:53Z

显示更改

Lr：/* patent */

2014-11-24T05:20:16Z

‎patent

Lr：/* QA */

2014-11-24T05:19:56Z

‎QA

Lr：/* QA */

2014-11-24T05:19:00Z

‎QA

Lr：/* Word2Vector */

2014-11-24T05:17:16Z

‎Word2Vector

@@ 第155行： / 第155行： @@
 ====Domain specific LM====
-* domain lm
+* domain lm(need to discuss with xiaoxi)
-:* embedded language model done
+:* embedded language model('''this week''')
-:* train some more LMs with Zhenlong (dianzishu sogou bbs chosen),put result on cvss.
+:* train some more LMs with Zhenlong (dianzishu sogou bbs chosen)("need result").
-:* small count done and ready to merge it.
+:* keep on training sogou2T lm(14/16 on 3rd iteration).('''this week''')
 *  new dict.
-:* dongxu help hanzhenglong to set up the new dict_2.0.
+:* handover of this work to hanzhenglong, give a simple docuemnt('''this week''')
-:* hanzhenglong need to report the result everyday on cvss
 ====tag LM====
-* summary done
 * need to do
-:* tag Probability should test add the weight(hanzhenglong) and handover to hanzhenglong (hold)
+:* tag Probability should test add the weight(hanzhenglong) and handover to hanzhenglong ('''this week''')
-:* make a summary about tag-lm and '''journal paper'''(wxx and yuanb)('''this weeks''').
+:* make a summary about tag-lm and '''journal paper'''(wxx and yuanb)('''two weeks''').
 ====RNN LM====
 *rnn
 :* test wer RNNLM on Chinese data from jietong-data('''this week''')
-:* generate the ngram model from rnnlm and test the ppl with different size txt.[http://cslt.riit.tsinghua.edu.cn/mediawiki/index.php/Jt-chinese#sampling_data_from_rnnlm]
+:* check the rnnlm code about how to Initialize and update learning rate.
 *lstm+rnn
-:* check the lstm-rnnlm code about how to Initialize and update learning rate.(hold)
+:* check the lstm-rnnlm code about how to Initialize and update learning rate.
 ===Word2Vector===
@@ 第185行： / 第261行： @@
 ====Knowledge vector====
 * Knowledge vector started
-:* generate the structured data from wiki
+:* begin to code
 ====Character to wordr====
 * Character to word conversion(hold)
@@ 第197行： / 第273行： @@
 ===QA===
-deatil:
+deatil:[http://cslt.riit.tsinghua.edu.cn/mediawiki/index.php/Hulan-2014-11-06]
 ====Spell mistake====
 * retrain the ngram model('''caoli''')
@@ 第205行： / 第281行： @@
 :* using MERT-4 method to get good value of multi-feature.like IDF,NER,baidu_weight,keyword etc.('''liurong this month''')
 ====Multi-Scene Recognition====
-* done
+* handover to duxk('''this week''')
 ====XiaoI framework====
 * give a report about xiaoI framework
 * new inter will install SEMPRE
 ====patent====
-* done
+* GA-method improve the QA('''this week''')

@@ 第155行： / 第155行： @@
 ====Domain specific LM====
-* domain lm(need to discuss with xiaoxi)
+* domain lm
-:* embedded language model('''this week''')
+:* embedded language model done
-:* train some more LMs with Zhenlong (dianzishu sogou bbs chosen)("need result").
+:* train some more LMs with Zhenlong (dianzishu sogou bbs chosen),put result on cvss.
-:* keep on training sogou2T lm(14/16 on 3rd iteration).('''this week''')
+:* small count done and ready to merge it.
 *  new dict.
-:* handover of this work to hanzhenglong, give a simple docuemnt('''this week''')
+:* dongxu help hanzhenglong to set up the new dict_2.0.
 ====tag LM====
+* summary done
-* different weight [http://cslt.riit.tsinghua.edu.cn/cgi-bin/cvss/cvss_request.pl?account=lr&step=view_request&cvssid=304 2014-Nov-23,Monday]
-:*
-{| border="2px"
-|+ different weight
-|-
-! method
-! tag-jsgf  !! corpus !! weight !! wer !! ser !! add_wer
-|-
-! experiment 3
-| 500(490 less frequent and 10 unseen)||500|| 0.1 || 16.72 || 77.92 || -
-|-
-−
-| || || 0.3 || 15.42 || 71.25 || -
-|-
-−
-| || || 0.5 || 15.40 || 69.58 || -
-|-
-−
-| || || 0.7 || 15.28 || 68.75|| -
-|-
-−
-| || || 0.8 || 15.38 || 68.33|| -
-|-
-−
-| || || 1 || 15.98 || 69.17|| -
-|-
-−
-| || || 2 || 19.08|| 70.83|| -
-|-
-! experiment 4
-|100(90 less frequent and 10 unseen) ||100 || 0.008 || 15.28|| 69.58|| -
-|-
-−
-| || || 0.02 || 14.84|| 69.58|| -
-|-
-−
-| || || 0.05 || 15.11|| 69.58|| -
-|-
-−
-| || || 0.1 || 15.30|| 69.75|| -
-|-
-−
-| || || 0.3 || 16.01|| 70.42|| -
-|-
-! experiment 5
-|500 ||100 || 0.01 || 17.57|| 78.75|| -
-|-
-−
-| || || 0.05 || 16.84|| 77.08|| -
-|-
-−
-| || || 0.08 || 16.59|| 76.25|| -
-|-
-−
-| || || 0.15 || 16.76|| 75.42|| -
-|-
-! experiment 6
-| 1280|| 500|| 0.1 || 17.42|| 77.92|| -
-|-
-−
-| || || 0.5 || 15.20|| 69.17|| -
-|-
-−
-| || || 0.8 || 15.30|| 68.33|| -
-|-
-−
-| || || 1 || 15.69|| 69.58|| -
-|-
-|}
-:* conclusion:
-. compare experiment 3  with experiment 5:
-    same jsgf file, but the  tag number in corpus if different, we can find that when add
-  more tag to corpus, the optimal weight is larger.
-. compare experiment 3 with experiment 6:
-   same tag number in corpus, but different jsgf size, we can find that different jsgf size have the
-  same optimal weight.
 * need to do
-:* tag Probability should test add the weight(hanzhenglong) and handover to hanzhenglong ('''this week''')
+:* tag Probability should test add the weight(hanzhenglong) and handover to hanzhenglong (hold)
-:* make a summary about tag-lm and '''journal paper'''(wxx and yuanb)('''two weeks''').
+:* make a summary about tag-lm and '''journal paper'''(wxx and yuanb)('''this weeks''').
 ====RNN LM====
 *rnn
 :* test wer RNNLM on Chinese data from jietong-data('''this week''')
-:* check the rnnlm code about how to Initialize and update learning rate.
+:* generate the ngram model from rnnlm and test the ppl with different size txt.[http://cslt.riit.tsinghua.edu.cn/mediawiki/index.php/Jt-chinese#sampling_data_from_rnnlm]
-:* generate the ngram model from rnnlm and test the ppl with different size txt.('''this week''')
 *lstm+rnn
-:* check the lstm-rnnlm code about how to Initialize and update learning rate.
+:* check the lstm-rnnlm code about how to Initialize and update learning rate.(hold)
 ===Word2Vector===
@@ 第261行： / 第185行： @@
 ====Knowledge vector====
 * Knowledge vector started
-:* begin to code
+:* generate the structured data from wiki
 ====Character to wordr====
 * Character to word conversion(hold)
@@ 第273行： / 第197行： @@
 ===QA===
-deatil:[http://cslt.riit.tsinghua.edu.cn/mediawiki/index.php/Hulan-2014-11-06]
+deatil:
 ====Spell mistake====
 * retrain the ngram model('''caoli''')
@@ 第281行： / 第205行： @@
 :* using MERT-4 method to get good value of multi-feature.like IDF,NER,baidu_weight,keyword etc.('''liurong this month''')
 ====Multi-Scene Recognition====
-* handover to duxk('''this week''')
+* done
 ====XiaoI framework====
 * give a report about xiaoI framework
 * new inter will install SEMPRE
 ====patent====
-* GA-method improve the QA('''this week''')
+* done

@@ 第91行： / 第91行： @@
    ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-===MSE===
+====MSE====
   (1) AURORA4 (train_clean)
      drop-retention/testcase(WER) | test_clean_wv1  | test_airport_wv1 | test_babble_wv1 | test_car_wv1
@@ 第100行： / 第100行： @@
      ---------------------------------------------------------------------------------------------------------
-===DAE(Deep Atuo-Encode)===
+====DAE(Deep Atuo-Encode)====
    (1) train_clean
      drop-retention/testcase(WER)| test_clean_wv1  | test_airport_wv1 | test_babble_wv1 | test_car_wv1

@@ 第64行： / 第64行： @@
 * Convolutive network (+)
 :* AURORA 4
   -----------------------------------------------------------------------------------------------------------------------
                    |  wer | hid-layers | hid-dim | delta-order | splice | lda-dim | learn-rate	| pooling | TBA
@@ 第79行： / 第80行： @@
    cnn_std_1200_3  | 6.66 |     5      | 1200	|      0      |    4   |   198   |   0.008	|   3     |patch-dim1 7
   -----------------------------------------------------------------------------------------------------------------------
-   ---------------------------------------------------------------------------------------------------------------------------------------------
+ :** 2)
-				| %WER		| Dnnhiddenlayers	| hid-dim	| pooling	| CNN_unit	|cnn_init_opts
+   ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-   ---------------------------------------------------------------------------------------------------------------------------------------------
+                                        | %WER          | Dnnhiddenlayers       | hid-dim       | pooling       | CNN_unit      |cnn_init_opts
    cnn_nonlda_std			| 5.73		| 4			| 1200    	| 3 		|  		|"--patch-dim1 8" input_dim ~ patch-dim1
-   ---------------------------------------------------------------------------------------------------------------------------------------------
+   ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
    cnn_nonlda_cnnunit_384		| 5.85		| 4			| 1200    	| 3		| 384		|"--patch-dim1 8 --num-filters2 384"
-   ---------------------------------------------------------------------------------------------------------------------------------------------
+   ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
    cnn_nonlda_cnnunit_220		| ----------    | 4			| 1200    	| 3		| 220		|"--patch-dim1 8 --num-filters2 220"
-   ---------------------------------------------------------------------------------------------------------------------------------------------
+   ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
 ===MSE===

@@ 第48行： / 第48行： @@
 * P-norm
+    ---------------------------------------------------------------------------------------------------------
           model/testcase(WER)    | test_clean_wv1  | test_airport_wv1 | test_babble_wv1 | test_car_wv1
      ---------------------------------------------------------------------------------------------------------
@@ 第64行： / 第64行： @@
 * Convolutive network (+)
 :* AURORA 4
                    |  wer | hid-layers | hid-dim | delta-order | splice | lda-dim | learn-rate	| pooling | TBA
   -----------------------------------------------------------------------------------------------------------------------
@@ 第78行： / 第79行： @@
    cnn_std_1200_3  | 6.66 |     5      | 1200	|      0      |    4   |   198   |   0.008	|   3     |patch-dim1 7
   -----------------------------------------------------------------------------------------------------------------------
+  ---------------------------------------------------------------------------------------------------------------------------------------------
 ---------------------------------------------------------------------------------------------------------------------------------------------
 				| %WER		| Dnnhiddenlayers	| hid-dim	| pooling	| CNN_unit	|cnn_init_opts
----------------------------------------------------------------------------------------------------------------------------------------------
+  ---------------------------------------------------------------------------------------------------------------------------------------------
-cnn_nonlda_std			| 5.73		| 4			| 1200    	| 3 		|  		|"--patch-dim1 8" input_dim ~ patch-dim1
+  cnn_nonlda_std			| 5.73		| 4			| 1200    	| 3 		|  		|"--patch-dim1 8" input_dim ~ patch-dim1
----------------------------------------------------------------------------------------------------------------------------------------------
+  ---------------------------------------------------------------------------------------------------------------------------------------------
-cnn_nonlda_cnnunit_384		| 5.85		| 4			| 1200    	| 3		| 384		|"--patch-dim1 8 --num-filters2 384"
+  cnn_nonlda_cnnunit_384		| 5.85		| 4			| 1200    	| 3		| 384		|"--patch-dim1 8 --num-filters2 384"
----------------------------------------------------------------------------------------------------------------------------------------------
+  ---------------------------------------------------------------------------------------------------------------------------------------------
-cnn_nonlda_cnnunit_220		| ----------    | 4			| 1200    	| 3		| 220		|"--patch-dim1 8 --num-filters2 220"
+  cnn_nonlda_cnnunit_220		| ----------    | 4			| 1200    	| 3		| 220		|"--patch-dim1 8 --num-filters2 220"
----------------------------------------------------------------------------------------------------------------------------------------------
+  ---------------------------------------------------------------------------------------------------------------------------------------------
-* MSE
+===MSE===
   (1) AURORA4 (train_clean)
      drop-retention/testcase(WER) | test_clean_wv1  | test_airport_wv1 | test_babble_wv1 | test_car_wv1
@@ 第98行： / 第98行： @@
      ---------------------------------------------------------------------------------------------------------
-* DAE(Deep Atuo-Encode)
+===DAE(Deep Atuo-Encode)===
    (1) train_clean
      drop-retention/testcase(WER)| test_clean_wv1  | test_airport_wv1 | test_babble_wv1 | test_car_wv1

←上一版本		2014年11月24日 (一) 05:19的版本
第331行：		第331行：
	* give a report about xiaoI framework		* give a report about xiaoI framework
	* new inter will install SEMPRE		* new inter will install SEMPRE
		+	====patent====
		+	* GA-method improve the QA