emnlp emnlp2010 emnlp2010-43 knowledge-graph by maker-knowledge-mining

43 emnlp-2010-Enhancing Domain Portability of Chinese Segmentation Model Using Chi-Square Statistics and Bootstrapping

Source: pdf

Author: Baobao Chang ; Dongxu Han

Abstract: Almost all Chinese language processing tasks involve word segmentation of the language input as their first steps, thus robust and reliable segmentation techniques are always required to make sure those tasks wellperformed. In recent years, machine learning and sequence labeling models such as Conditional Random Fields (CRFs) are often used in segmenting Chinese texts. Compared with traditional lexicon-driven models, machine learned models achieve higher F-measure scores. But machine learned models heavily depend on training materials. Although they can effectively process texts from the same domain as the training texts, they perform relatively poorly when texts from new domains are to be processed. In this paper, we propose to use χ2 statistics when training an SVM-HMM based segmentation model to im- prove its ability to recall OOV words and then use bootstrapping strategies to maintain its ability to recall IV words. Experiments show the approach proposed in this paper enhances the domain portability of the Chinese word segmentation model and prevents drastic decline in performance when processing texts across domains.

Reference: text

Summary: the most important sentenses genereted by tfidf model

sentIndex sentText sentNum sentScore

1 edu Abstract Almost all Chinese language processing tasks involve word segmentation of the language input as their first steps, thus robust and reliable segmentation techniques are always required to make sure those tasks wellperformed. [sent-4, score-0.766]

2 Although they can effectively process texts from the same domain as the training texts, they perform relatively poorly when texts from new domains are to be processed. [sent-8, score-0.389]

3 In this paper, we propose to use χ2 statistics when training an SVM-HMM based segmentation model to im- prove its ability to recall OOV words and then use bootstrapping strategies to maintain its ability to recall IV words. [sent-9, score-0.97]

4 Experiments show the approach proposed in this paper enhances the domain portability of the Chinese word segmentation model and prevents drastic decline in performance when processing texts across domains. [sent-10, score-0.684]

5 95 can be achieved in the closed test tracks, in which only specified training materials can be used in learning segmentation models1. [sent-15, score-0.524]

6 Traditional word segmentation approaches are lexicon-driven (Liang, 1987) and assume predefined lexicons of Chinese words are available. [sent-16, score-0.445]

7 Xue (2003) proposed a novel way of segmenting Chinese texts, and it views the Chinese word segmentation task as a character tagging task. [sent-19, score-0.547]

8 According to Xue’s approach, no predefined Chinese lexicons are required; a tagging model is learned by using manually segmented training texts. [sent-20, score-0.239]

9 The model is then used to assign each character a tag indicating the position of this character within a word. [sent-21, score-0.293]

10 Xue’s approach has become the most popular approach to Chinese word segmentation for its high performance and unified way of dealing with out-of-vocabulary (OOV) issues. [sent-22, score-0.383]

11 Most segmentation work began to follow this approach later. [sent-23, score-0.383]

12 3) New feature templates were added, such as the templates that were used in representing numbers, dates, letters etc. [sent-35, score-0.467]

13 , 2005) Character tagging approaches require manually segmented training texts to learn models usually in a supervised way. [sent-37, score-0.289]

14 The performance is always evaluated on a test set from the same domain as the training set. [sent-38, score-0.198]

15 Actually, when test set is from other domains than the domain where training set is from, the learned model normally underperforms substantially. [sent-40, score-0.302]

16 Actually, even when the test set has the same domain properties as the training set, the ability of the model to recall OOV words is still the main obstacle to achieve better performance of segmentation. [sent-42, score-0.403]

17 However, when the test set is different with the training set in nature, the OOV recall normally drops much more substantially, and becomes much lower. [sent-43, score-0.237]

18 However, the approach did not show a segmentation performance as good as that of the supervised approach. [sent-49, score-0.383]

19 Aiming at preventing the OOV recall from dropping sharply and still maintaining an overall performance as good as that of the state-of-art segmenter when working with heterogeneous test sets, we propose in this paper to use a semisupervised way for Chinese word segmentation task. [sent-53, score-0.583]

20 Specifically, we propose to use χ2 statistics together with bootstrapping strategies to build Chinese word segmentation model. [sent-54, score-0.62]

21 Our work differs from the previous supervised work in its ability to cope with domain variation and differs from the previous unsupervised work in its much better overall segmentation performance. [sent-59, score-0.625]

22 The rest of the paper is organized as follows: In section 2, we give a brief introduction to the hidden Markov Support Vector Machines, on which we rely to build the segmentation model. [sent-60, score-0.429]

23 In section 3, we list the segmentation tags and the basic feature templates we used in the paper. [sent-61, score-0.582]

24 2 The hidden Markov support vector machines The hidden Markov support vector machine (SVM-HMM) is actually a special case of the structural support vector machines proposed by Tsochantaridis et al. [sent-65, score-0.472]

25 It differs from support vector machine in its ability to model complex structured problems and shares the max-margin training principles with support vector machines. [sent-68, score-0.328]

26 The hidden Markov support vector machine model is inspired by the hidden Markov model and is an instance of structural support vector machine dedicated to solve sequence labeling learning, a problem that CRF model is assumed to solve. [sent-69, score-0.499]

27 f(x; w) = arg max F(x, y;w) y∈Y In the structural SVMs, F is assumed to be linear 790 in some combined feature representation of the input sequence and the label sequence ψ(x,y), i. [sent-71, score-0.196]

28 , δ(yt,y)=⎧⎨⎩10 i f y t =≠yy The first term of the discriminant function is used to model the interactions between input features and labels, and the second term is used to model interactions between nearby labels. [sent-80, score-0.282]

29 We use the following feature templates, as are widely used in most segmentation work: (a) Cn (n = -2, -1, 0, 1, 2) (b) CnCn+1 (n = -2, -1, 0, 1) (c) C-1C+1 791 Here C refers to a character; n refers to the position index relative to the current character. [sent-88, score-0.519]

30 By setting the above feature templates, we actually set a 5character window to extract features, the current character, 2 characters to its left and 2 characters to its right. [sent-89, score-0.216]

31 In addition, we also use the following feature templates to extract features representing the character type: (d) Tn (n = -2, -1, 0, 1, 2) (e) TnTn+1 (n = -2, -1, 0, 1) (f) T-1T+1 Here T refers to a character type, and its value can be digit, letter, punctuation or Chinese character. [sent-90, score-0.533]

32 The type feature is important, for there are two versions of Arabic numbers, Latin alphabets and punctuations in the Chinese texts. [sent-91, score-0.205]

33 If the type features are not introduced, most of the numbers, letters and punctuation of single-byte can not be segmented correctly. [sent-96, score-0.247]

34 The type feature establishes links between the two versions of codes, for both versions of a digit, a letter or punctuation share the same type feature value. [sent-97, score-0.3]

35 That is the mapping all singlebyte versions of digits, letters and punctuations in the test sets into their double-byte counterparts as in the training set. [sent-99, score-0.396]

36 4 The χ2 statistic features χ2 test is one of hypothesis test methods, which can be used to test if two events co-occur just by chance or not. [sent-101, score-0.269]

37 Unsupervised segmentation approach also mainly relies on mutual information and t-score to identify words in Chinese texts (Sun et al. [sent-104, score-0.5]

38 Inspired by their work, we believe that χ2 statistics could also be incorporated into supervised segmentation models to deal with the OOV issue. [sent-106, score-0.491]

39 If two continuous characters in the test set have a higher χ2 score, it is highly likely they form a word or are part of a word even they are not seen in the training set. [sent-108, score-0.21]

40 + + + χn2orm(C1,C2)=⎢⎣⎢χ2(χCm21a,xC−2)χ−m2inχm2in×10⎥⎦⎥ To make the learned model sensitive to the χ2 statistics, we then add two more feature templates as follows: (g) XnXn+1 (n = -2, -1, 0, 1) (h) X-1X+1 The value of the feature XnXn+1 is the normalized χ2 score of the bigram CnCn+1. [sent-116, score-0.353]

41 By using the χ2 features, statistics from the test set are introduced into segmentation model, and it makes the resulted model more aware of the test set and therefore more robust to test domains other than training domains. [sent-118, score-0.781]

42 All features generated from the templates (a)-(f) together with the 55 χ2 features form the whole … … …, 792 feature set. [sent-120, score-0.199]

43 The feature representation of the training set is then used to learn the model and the feature representation of the test set is then used for segmentation and evaluated by comparison with gold standard segmentation. [sent-122, score-0.606]

44 By this way, an OOV word in the test set might be found by the segmentation model if the bigrams extracted from this word take higher χ2 scores. [sent-124, score-0.556]

45 5 the bootstrapping strategy The addition of the χ2 features can be also problematic as we will see in the experiments. [sent-125, score-0.192]

46 To keep the IV recall from falling, we propose to use a bootstrapping method. [sent-130, score-0.255]

47 Then we do the segmentation for the test set with the two models simultaneously. [sent-133, score-0.494]

48 It is not difficult to understand that the intersection of the two results has both high OOV recall and high IV recall, if we also extract these sentences from the gold standard segmentation and perform evaluations. [sent-138, score-0.584]

49 We then put the intersection results into the training set to form a new training set. [sent-139, score-0.197]

50 1 On the training set and test set For training the segmentation model, we use the training data provided by Peking University for 793 bakeoff . [sent-150, score-0.594]

51 The PKU training data is actually consisted of all texts of the People’s Daily newspaper in January of 1998. [sent-153, score-0.19]

52 A characteristic of the PKU data is that all Arabic numbers, Latin letters and punctuations in the data are all double-byte GB codes; there are no single-byte ASCII versions of these 20052 characters in the PKU training data. [sent-155, score-0.336]

53 We use this test set to show how well the SVM-HMM can be used to model segmentation problem and the performance that a segmentation model achieves when applied to the texts from the same domain. [sent-159, score-1.034]

54 The second and the third test sets are set to test how well the segmentation model can apply to texts from other domains. [sent-160, score-0.697]

55 In addition, we also evaluate all the test results with sentence accuracies (SA), which is the proportion of the correctly segmented sentences in the test set. [sent-173, score-0.238]

56 CRF To show how well the SVM-HMM model can be used to model segmentation tasks and its performance compared to that of CRF model, we use the training set to train two models, one with SVMHMM and the other with CRF. [sent-216, score-0.514]

57 However, SVM-HMM takes interactions between nearby labels into the model, which means there is a label bigram feature template implicitly used in the SVM-HMM. [sent-219, score-0.259]

58 The performances of the two models are shown in Table-2, which shows SVM-HMM can be used to model Chinese segmentation tasks 6 specified by the B template as the toolkit requires. [sent-236, score-0.517]

59 2 The baseline model To test how well the segmentation model applies to other domain texts, we only use the SVM-HMM model with the same parameters as in section 6. [sent-239, score-0.655]

60 For a baseline model, we only use feature templates (a)-(c), the performances of the basic model on the three test sets are shown in Table-3. [sent-241, score-0.315]

61 For the test set A, which is from the same domain as the training data, an F-score 0. [sent-242, score-0.198]

62 Especially the OOV recalls fall drastically, which means the model is very sensitive to the domain variation. [sent-245, score-0.426]

63 This also shows the domain portability of the segmentation model is still an obstacle for the segmentation model to be used in crossdomain applications. [sent-247, score-1.013]

64 Especially, test set C is full of single- byte version of such numbers, letters and punctuations. [sent-250, score-0.188]

65 Therefore, we use the feature tem-plates (a)-(f) to train a type sensitive model with the training data. [sent-252, score-0.2]

66 The IV recall even has a slight fall while the OOV recall rises a little. [sent-255, score-0.3]

67 Different with the test set A, even the IV recall for test set C rises slightly. [sent-262, score-0.329]

68 Unlike test set C, there are not so many singlebyte characters in test set B. [sent-264, score-0.277]

69 Even though the OOV 795 recall does rise significantly, the change in OOV recall for test set B is not as much as that for test set B. [sent-265, score-0.35]

70 4 The χ2-based model Compared with OOV recall for test set A, the OOV recall for test set B and C are still lower. [sent-268, score-0.385]

71 To promote the OOV recall, we use the feature templates (a)-(h) to train a χ2-based model with the training data. [sent-269, score-0.362]

72 But the OOV recall for the three test sets does improve, especially for test set B and C. [sent-273, score-0.256]

73 The IV recalls for the three test sets drop, especially for test set B and C. [sent-274, score-0.399]

74 5 Bootstrapping To increase the OOV recall and prevent the IV recall from falling, we use the bootstrapping strategy in section 5. [sent-277, score-0.38]

75 We just do the bootstrapping to test set B and C, because what we are concerned with in this paper is to improve the performance of the model to different domains. [sent-279, score-0.277]

76 To illustrate how the bootstrapping strategy works, we also present the performance of the intermediate models on test set C in each pass of the bootstrapping in table-7 and table-8. [sent-283, score-0.551]

77 Table-7 is results of the intermediate χ2-based models for test set C. [sent-284, score-0.198]

78 Table-8 is results of the intermediate nonχ2-based models for test set C. [sent-285, score-0.198]

79 Figure-2 illustrates changes in OOV recalls of both non- χ2-based models and χ2-based models as the bootstrapping algorithm advances for test set C. [sent-286, score-0.631]

80 Figure-3 illustrates changes in IV re-calls of both non- χ2-based models and χ2-based models for test set C. [sent-287, score-0.204]

81 Performance of the intermediate χ2-based models for test set C I0312 0 . [sent-289, score-0.198]

82 Performance of the intermediate for test set C non-χ2-based models 10432I 0. [sent-296, score-0.198]

83 Performance of the intersection of the intermediate model and non-χ2-based χ2-based model for test C I0312 0 . [sent-303, score-0.345]

84 65S796A78912364 recall of the χ2-based model as the bootstrapping algorithm advances. [sent-309, score-0.29]

85 The abilities to recall IV words of both models improve, and even the final IV recall of the χ2-based model surpasses the IV recall of the type sensitive model shown in Table-3. [sent-310, score-0.448]

86 We just show the changes in OOV recalls and IV recalls as illustrated in Figure-4 and Figure5. [sent-316, score-0.537]

87 One can see from Figure-4 and Figure-5, the bootstrapping strategy also works for test set B in a similar way as it works for test set C. [sent-317, score-0.354]

88 5 657801234 without chi-square featureswith chi-square features Figure-2 the Changes in OOV recalls for test set C as boot-strapping algorithm advances 796 0 0 0 . [sent-322, score-0.418]

89 3195294 9 053 5 5421501234 without chi-square featureswith chi-square features Figure-3 the Changes in IV recalls for test set C as boot-strapping algorithm advances 0 0 0. [sent-327, score-0.418]

90 However, as the algorithm progresses, both the OOV recall and IV recall of the intersection results fall, but are still higher than OOV recall and IV recall of the final results on the whole test set. [sent-333, score-0.564]

91 As we said before, we give also sentence accuracies of all segmentation models. [sent-334, score-0.383]

92 With the χ2 statistics and bootstrapping strategies, the sentence accuracy also rises. [sent-335, score-0.212]

93 7 Conclusions Sequence labeling models are widely used in Chinese word segmentation recently. [sent-339, score-0.46]

94 High performance can be achieved when the test data is from the same domain as the training data. [sent-340, score-0.198]

95 However, if the test data is assumed to be from other domains than the domain of the training data, the segmentation models always underperform substantially. [sent-341, score-0.675]

96 To enhance the portability of the sequence labeling segmentation models to other domains, this paper proposes to use χ2 statistics and bootstrapping strategy. [sent-342, score-0.788]

97 The experiment shows the approach significantly increases both IV recall and OOV recall when processing texts from different domains. [sent-343, score-0.305]

98 We also show in this paper that hidden Markov support vector machine which is also a sequence labeling model like CRF can be used to model the Chinese word segmentation problem, by which 797 high F-score results can be obtained like those of CRF model. [sent-344, score-0.687]

99 Duan Huiming for her work in preparing the gold standard segmentation and to the anonymous reviewers for their comments to the paper. [sent-352, score-0.383]

100 , 2004, Chinese word segmentation without using dictionary based on unsupervised learning strategy. [sent-374, score-0.383]

similar papers computed by tfidf model

tfidf for this paper:

wordName wordTfidf (topN-words)

[('oov', 0.498), ('segmentation', 0.383), ('iv', 0.336), ('recalls', 0.237), ('chinese', 0.207), ('bootstrapping', 0.161), ('templates', 0.161), ('crf', 0.133), ('character', 0.129), ('texts', 0.117), ('letters', 0.107), ('intersection', 0.107), ('peking', 0.101), ('recall', 0.094), ('intermediate', 0.087), ('domain', 0.086), ('pku', 0.081), ('punctuations', 0.081), ('test', 0.081), ('segmented', 0.076), ('rises', 0.073), ('featureswith', 0.071), ('characters', 0.068), ('promote', 0.067), ('portability', 0.063), ('changes', 0.063), ('discriminant', 0.061), ('sighan', 0.059), ('bigrams', 0.057), ('xue', 0.055), ('nearby', 0.055), ('latin', 0.055), ('ascii', 0.055), ('sequence', 0.053), ('markov', 0.053), ('bigram', 0.052), ('statistics', 0.051), ('versions', 0.049), ('vector', 0.049), ('refers', 0.049), ('ability', 0.048), ('interactions', 0.048), ('labeling', 0.047), ('cncn', 0.047), ('singlebyte', 0.047), ('xnxn', 0.047), ('hidden', 0.046), ('china', 0.044), ('actually', 0.042), ('codes', 0.042), ('numbers', 0.041), ('digit', 0.041), ('falling', 0.041), ('island', 0.041), ('jeju', 0.041), ('tsochantaridis', 0.041), ('template', 0.04), ('support', 0.039), ('fall', 0.039), ('domains', 0.038), ('differs', 0.038), ('feature', 0.038), ('type', 0.037), ('machines', 0.037), ('bakeoff', 0.037), ('tagging', 0.035), ('model', 0.035), ('sun', 0.034), ('yt', 0.034), ('predefined', 0.032), ('cutoff', 0.032), ('cope', 0.032), ('digits', 0.032), ('normally', 0.031), ('arabic', 0.031), ('strategy', 0.031), ('training', 0.031), ('lexicons', 0.03), ('train', 0.03), ('continuous', 0.03), ('daily', 0.03), ('models', 0.03), ('advances', 0.029), ('specified', 0.029), ('sensitive', 0.029), ('obstacle', 0.028), ('put', 0.028), ('encoding', 0.028), ('punctuation', 0.027), ('dc', 0.027), ('margin', 0.027), ('deal', 0.027), ('label', 0.026), ('assumed', 0.026), ('drop', 0.026), ('statistic', 0.026), ('xt', 0.026), ('strategies', 0.025), ('letter', 0.025), ('segmenter', 0.025)]

similar papers list:

simIndex simValue paperId paperTitle

same-paper 1 0.99999964 43 emnlp-2010-Enhancing Domain Portability of Chinese Segmentation Model Using Chi-Square Statistics and Bootstrapping

Author: Baobao Chang ; Dongxu Han

2 0.24406253 69 emnlp-2010-Joint Training and Decoding Using Virtual Nodes for Cascaded Segmentation and Tagging Tasks

Author: Xian Qian ; Qi Zhang ; Yaqian Zhou ; Xuanjing Huang ; Lide Wu

Abstract: Many sequence labeling tasks in NLP require solving a cascade of segmentation and tagging subtasks, such as Chinese POS tagging, named entity recognition, and so on. Traditional pipeline approaches usually suffer from error propagation. Joint training/decoding in the cross-product state space could cause too many parameters and high inference complexity. In this paper, we present a novel method which integrates graph structures of two subtasks into one using virtual nodes, and performs joint training and decoding in the factorized state space. Experimental evaluations on CoNLL 2000 shallow parsing data set and Fourth SIGHAN Bakeoff CTB POS tagging data set demonstrate the superiority of our method over cross-product, pipeline and candidate reranking approaches.

3 0.18792263 2 emnlp-2010-A Fast Decoder for Joint Word Segmentation and POS-Tagging Using a Single Discriminative Model

Author: Yue Zhang ; Stephen Clark

Abstract: We show that the standard beam-search algorithm can be used as an efficient decoder for the global linear model of Zhang and Clark (2008) for joint word segmentation and POS-tagging, achieving a significant speed improvement. Such decoding is enabled by: (1) separating full word features from partial word features so that feature templates can be instantiated incrementally, according to whether the current character is separated or appended; (2) deciding the POS-tag of a potential word when its first character is processed. Early-update is used with perceptron training so that the linear model gives a high score to a correct partial candidate as well as a full output. Effective scoring of partial structures allows the decoder to give high accuracy with a small beam-size of 16. In our 10-fold crossvalidation experiments with the Chinese Tree- . bank, our system performed over 10 times as fast as Zhang and Clark (2008) with little accuracy loss. The accuracy of our system on the standard CTB 5 test was competitive with the best in the literature. 1 Introduction and Motivation Several approaches have been proposed to solve word segmentation and POS-tagging jointly, including the reranking approach (Shi and Wang, 2007; Jiang et al., 2008b), the hybrid approach (Nakagawa and Uchimoto, 2007; Jiang et al., 2008a), and the single-model approach (Ng and Low, 2004; Zhang and Clark, 2008; Kruengkrai et al., 2009). These methods led to accuracy improvements over the traditional, pipelined segmentation and POS-tagging . . . 843 clark} @ cl cam ac uk baseline by avoiding segmentation error propagation and making use of part-of-speech information to improve segmentation. The single-model approach to joint segmentation and POS-tagging offers consistent training of all in- formation, concerning words, characters and partsof-speech. However, exact inference with dynamic programming can be infeasible if features are defined over a large enough range of the output, such as over a two-word history. In our previous work (Zhang and Clark, 2008), which we refer to as Z&C08; from now on, we used an approximate decoding algorithm that keeps track of a set of partially built structures for each character, which can be seen as a dynamic programming chart which is greatly reduced by pruning. In this paper we follow the line of single-model research, in particular the global linear model of Z&C08.; We show that effective decoding can be achieved with standard beam-search, which gives significant speed improvements compared to the decoding algorithm of Z&C08;, and achieves accuracies that are competitive with the state-of-the-art. Our research is also in line with recent research on improving the speed of NLP systems with little or no accuracy loss (Charniak et al., 2006; Roark and Hollingshead, 2008). Our speed improvement is achieved by the use of a single-beam decoder. Given an input sentence, candidate outputs are built incrementally, one character at a time. When each character is processed, it is combined with existing candidates in all possible ways to generate new candidates, and an agenda is used to keep the N-best candidate outputs from ProceMedITin,g Ms oasfs thaceh 2u0se1t0ts C,o UnSfAer,e n9c-e1 on O Ectmobpeir ic 2a0l1 M0.e ?tc ho2d0s10 in A Nsastoucira tlio Lnan fogru Cagoem Ppruotcaetisosninagl, L pinag eusis 8t4ic3s–852, the begining of the sentence to the current character. Compared to the multiple-beam search algorithm of Z&C08;, the use of a single beam can lead to an order of magnitude faster decoding speed. 1.1 The processing of partial words An important problem that we solve in this paper is the handling of partial words with a single beam decoder for the global model. As we pointed out in Z&C08;, it is very difficult to score partial words properly when they are compared with full words, although such comparison is necessary for incremental decoding with a single-beam. To allow comparisons with full words, partial words can either be treated as full words, or handled differently. We showed in Z&C08; that a naive single-beam decoder which treats partial words in the same way as full words failed to give a competitive accu- racy. An important reason for the low accuracy is over-segmentation during beam-search. Consider the three characters “ 自来水 (tap water)”. The first two characters do not make sense when put together as a single word. Rather, when treated as two singlecharacter words, they can make sense in a sentence such as “请 (please) 自 (self) 来 (come) 取 (take)”. Therefore, when using single-beam search to process “ 自来水 (tap water)”, the two-character word candidate “ 自来” is likely to have been thrown off the agenda before the third character “水” is considered, leading to an unrecoverable segmentation error. This problem is even more severe for a joint segmentor and POS-tagger than for a pure word segmentor, since the POS-tags and POS-tag bigram of “ 自 and “来” further supports them being separated when ”来” is considered. The multiple-beam search decoder we proposed in Z&C08; can be seen as a means to ensure that the three characters “ 自来水” always have a chance to be considered as a single word. It explores candidate segmentations from the beginning of the sentence until each character, and avoids the problem of processing partial words by considering only full words. However, since it ex- ” plores a larger part of the search space than a singlebeam decoder, its time complexity is correspondingly higher. In this paper, we treat partial words differently from full words, so that in the previous example, 844 the decoder can take the first two characters in “ 自来水 (tap water)” as a partial word, and keep it in the beam before the third character is processed. One challenge is the representation of POS-tags for partial words. The POS of a partial word is undefined without the corresponding full word information. Though a partial word can make sense with a particular POS-tag when it is treated as a complete word, this POS-tag is not necessarily the POS of the full word which contains the partial word. Take the three-character sequence “下雨天” as an example. The first character “下” represents a singlecharacter word “below”, for which the POS can be LC or VV. The first two characters “下雨” represent a two-character word “rain”, for which the POS can be VV. Moreover, all three characters when put together make the word “rainy day”, for which the POS is NN. As discussed above, assigning POS tags to partial words as if they were full words leads to low accuracy. An obvious solution to the above problem is not to assign a POS to a partial word until it becomes a full word. However, lack of POS information for partial words makes them less competitive compared to full words in the beam, since the scores of full words are futher supported by POS and POS ngram information. Therefore, not assigning POS to partial words potentially leads to over segmentation. In our experiments, this method did not give comparable accuracies to our Z&C08; system. In this paper, we take a different approach, and assign a POS-tag to a partial word when its first character is separated from the final character of the previous word. When more characters are appended to a partial word, the POS is not changed. The idea is to use the POS of a partial word as the predicted POS of the full word it will become. Possible predictions are made with the first character of the word, and the likely ones will be kept in the beam for the next processing steps. For example, with the three characters “下雨天”, we try to keep two partial words (besides full words) in the beam when the first word “下” is processed, with the POS being VV and NN, respectively. The first POS predicts the two-character word “下雨” ， and the second the three-character word “下雨天”. Now when the second character is processed, we still need to maintain the possible POS NN in the agenda, which predicts the three-character word “下雨天”. As a main contribution of this paper, we show that the mechanism ofpredicting the POS at the first character gives competitive accuracy. This mechanism can be justified theoretically. Unlike alphabetical languages, each Chinese character represents some specific meanings. Given a character, it is natural for a human speaker to know immediately what types of words it can start. The allows the knowledge of possible POS-tags of words that a character can start, using information about the character from the training data. Moreover, the POS of the previous words to the current word are also useful in deciding possible POS for the word.1 The mechanism of first-character decision of POS also boosts the efficiency, since the enumeration of POS is unecessary when a character is appended to the end of an existing word. As a result, the complexity of each processing step is reduce by half compared to a method without POS prediction. Finally, an intuitive way to represent the status of a partial word is using a flag explicitly, which means an early decision of the segmentation of the next incoming character. We take a simpler alternative approach, and treat every word as a partial word until the next incoming character is separated from the last character of this word. Before a word is confirmed as a full word, we only apply to it features that represent its current partial status, such as character bigrams, its starting character and its part-ofspeech, etc. Full word features, including the first and last characters of a word, are applied immediately after a word is confirmed as complete. An important component for our proposed system is the training process, which needs to ensure that the model scores a partial word with predicted POS properly. We use the averaged perceptron (Collins, 2002) for training, together with the “early update” mechanism of Collins and Roark (2004). Rather than updating the parameters after decoding is com- plete, the modified algorithm updates parameters at any processing step if the correct partial candidate falls out of the beam. In our experiments using the Chinese Treebank 1The next incoming characters are also a useful source of information for predicting the POS. However, our system achieved competitive accuracy with Z&C08; without such character lookahead features. 845 data, our system ran an order of magnitude faster than our Z&C08; system with little loss of accuracy. The accuracy of our system was competitive with other recent models. 2 Model and Feature Templates We use a linear model to score both partial and full candidate outputs. Given an input x, the score of a candidate output y is computed as: Score(y) = Φ(y) · where Φ(y) is the global feature vector extracted from y, and is the parameter vector of the model. Figure 1 shows the feature templates for the model, where templates 1 14 contain only segmentation information and templates 15 29 contain w~ , w~ – – both segmentation and POS information. Each template is instantiated according to the current character in the decoding process. Row “For” shows the conditions for template instantiation, where “s” indicates that the corresponding template is instantiated when the current character starts a new word, and “a” indicates that the corresponding template is instantiated when the current character does not start a new word. In the row for feature templates, w, t and c are used to represent a word, a POS-tag and a character, respectively. The subscripts are based on the current character, where w−1 represents the first word to the left of the current character, and p−2 represents the POS-tag on the second word to the left of the current character, and so on. As an example, feature template 1is instantiated when the current character starts a new word, and the resulting feature value is the word to the left of this character. start(w), end(w) and len(w) represent the first character, the last character and the length of word w, respectively. The length of a word is normalized to 16 if it is larger than 16. cat(c) represents the POS category of character c, which is the set of POS-tags seen on character c, as we used in Z&C08.; Given a partial or complete candidate y, its global feature vector Φ(y) is computed by instantiating all applicable feature templates from Table 1 for each character in y, according to whether or not the character is separated from the previous character. The feature templates are mostly taken from, or inspired by, the feature templates of Z&C08.; Templates 1, 2, 3, 4, 5, 8, 10, 12, 13, 14, 15, 19, 20, Feature templateFor 24, 27 and 29 concern complete word information, and they are used in the model to differentiate correct and incorrect output structures in the same way as our Z&C08; model. Templates 6, 7, 9, 16, 17, 18, 21, 22, 23, 25, 26 and 28 concern partial word information, whose role in the model is to indicate the likelihood that the partial word including the current character will become a correct full word. They act as guidance for the action to take for the cur846 function DECODE(sent, agenda): CLEAR(agenda) ADDITEM(agenda, “”) for index in [0..LEN(sent)]: for cand in agenda: new ← APPEND(cand, sent[index]) ADDITEM(agenda, new) for pos in TAGSET(): new ← SEP(cand, sent[index], pos) ADDITEM(agenda, new) agenda ← N-BEST(agenda) retaugrenn BEST(agenda) Figure 1: The incremental beam-search decoder. rent character according to the context, and are the crucial reason for the effectiveness of the algorithm with a small beam-size. 2.1 Decoding The decoding algorithm builds an output candidate incrementally, one character at a time. Each character can either be attached to the current word or separated as the start a new word. When the current character starts a new word, a POS-tag is assigned to the new word. An agenda is used by the decoder to keep the N-best candidates during the incremental process. Before decoding starts, the agenda is initialized with an empty sentence. When a character is processed, existing candidates are removed from the agenda and extended with the current character in all possible ways, and the N-best newly generated candidates are put back onto the agenda. After all input characters have been processed, the highest-scored candidate from the agenda is taken as the output. Pseudo code for the decoder is shown in Figure 1. CLEAR removes all items from the agenda, ADDITEM adds a new item onto the agenda, N-BEST returns the N highest-scored items from the agenda, and BEST returns the highest-scored item from the agenda. LEN returns the number of characters in a sentence, and sent[i] returns the ith character from the sentence. APPEND appends a character to the last word in a candidate, and SEP joins a character as the start of a new word in a candidate, assigning a POS-tag to the new word. Both our decoding algorithm and the decoding algorithm of Z&C08; run in linear time. However, in order to generate possible candidates for each character, Z&C08; uses an extra loop to search for possible words that end with the current character. A restriction to the maximum word length is applied to limit the number of iterations in this loop, without which the algorithm would have quadratic time complexity. In contrast, our decoder does not search backword for the possible starting character of any word. Segmentation ambiguities are resolved by binary choices between the actions append or separate for each character, and no POS enumeration is required when the character is appended. This improves the speed by a significant factor. 2.2 Training The learning algorithm is based on the generalized perceptron (Collins, 2002), but parameter adjustments can be performed at any character during the decoding process, using the “early update” mechanism of Collins and Roark (2004). The parameter vector of the model is initialized as all zeros before training, and used to decode training examples. Each training example is turned into the raw input format, and processed in the same way as decoding. After each character is processed, partial candidates in the agenda are compared to the corresponding gold-standard output for the same characters. If none of the candidates in the agenda are correct, the decoding is stopped and the parameter vector is updated by adding the global feature vector of the gold-standard partial output and subtracting the global feature vector of the highest-scored partial candidate in the agenda. The training process then moves on to the next example. However, if any item in the agenda is the same as the corresponding gold-standard, the decoding process moves to the next character, without any change to the parameter values. After all characters are processed, the decoder prediction is compared with the training example. If the prediction is correct, the parameter vector is not changed; otherwise it is updated by adding the global feature vector of the training example and subtracting the global feature vector of the decoder prediction, just as the perceptron algorithm does. The same training examples can be used to train the model for multiple iterations. We use 847 the averaged parameter vector (Collins, 2002) as the final model. Pseudocode for the training algorithm is shown in Figure 2. It is based on the decoding algorithm in Figure 1, and the main differences are: (1) the training algorithm takes the gold-standard output and the parameter vector as two additional arguments; (2) the training algorithm does not return a prediction, but modifies the parameter vector when necessary; (3) lines 11to 20 are additional lines of code for parameter updates. Without lines 11 to 16, the training algorithm is exactly the same as the generalized perceptron algorithm. These lines are added to ensure that the agenda contains highly probable candidates during the whole beam-search process, and they are crucial to the high accuracy of the system. As stated earlier, the decoder relies on proper scoring of partial words to maintain a set of high quality candidates in the agenda. Updating the value of the parameter vector for partial outputs can be seen as a means to ensure correct scoring of partial candidates at any character. 2.3 Pruning We follow Z&C08; and use several pruning methods, most of which serve to to improve the accuracy by removing irrelevant candidates from the beam. First, the system records the maximum number of characters that a word with a particular POS-tag can have. For example, from the Chinese Treebank that we used for our experiments, most POS are associated with only with one- or two-character words. The only POS-tags that are seen with words over ten characters long are NN (noun), NR (proper noun) and CD (numbers). The maximum word length information is initialized as all ones, and updated according to each training example before it is processed. Second, a tag dictionary is used to record POStags associated with each word. During decoding, frequent words and words with “closed set” tags2 are only allowed POS-tags according to the tag dictionary, while other words are allowed every POS-tag to make candidate outputs. Whether a word is a frequent word is decided by the number of times it has been seen in the training process. Denoting the num2“Closed set” tags are the set of POS-tags which are only associated with a fixed set of words, according to the Penn Chinese Treebank specifications (Xia, 2000). function TRAIN(sent, agenda, gold-standard, w~ ): 01: CLEAR(agenda) 02: ADDITEM(agenda, “”) 03: for index in [0..LEN(sent)]: 04: 05: 06: 07: 08: 09: 10: 11: 12: 13: 14: 15: 16: 17: 18: 19: 20: 21: for cand in agenda: new ← APPEND(cand, sent[index]) ADDITEM(agenda, new) for pos in TAGSET(): new ← SEP(cand, sent[index], pos) ADDITEM(agenda, new) agenda ← N-BEST(agenda) faogre cnadnad ← ←in agenda: if cand = gold-standard[0:index] : CONTINUE w~ ← w~ + Φ(gold-standard[0:index]) ww~~ ← ww~ ~ - Φ(BEST(agenda)) wr~et ←urn w~ if BEST(agenda) gold-standard: w~ ← a ~wg + Φ(gold-standard) ww~~ ← ww~ ~ - Φ(BEST(agenda)) wr~et ←urn w~ return = Figure 2: The incremental learning function. ber of times the most frequent word has been seen with M, a word is a frequent word if it has been seen more than M/5000 5 times. The threshold value is taken from Z&C08;, and we did not adjust it during development. Word frequencies are initialized as zeros and updated according to each training example before it is processed; the tag dictionary is initialized as empty and updated according to each training example before it is processed. Third, we make an additional record of the initial characters for words with “closed set” tags. During decoding, when the current character is added as the start of a new word, “closed set” tags are only assigned to the word if it is consistent with the record. This type of pruning is used in addition to the tag + dictionary to prune invalid partial words, while the tag dictionary is used to prune complete words. The record for initial character and POS is initially empty, and udpated according to each training example before it is processed. Finally, at any decoding step, we group partial 848 candidates that are generated by separating the current character as the start of a new word by the signature p0p−1w−1, and keep only the best among those having the same p0p−1w−1. The signature p0p−1w−1 is decided by the feature templates we use: it can be shown that if two candidates cand1 and cand2 generated at the same step have the same signature, and the score of cand1 is higher than the score of cand2, then at any future step, the highest scored candidate generated from cand1 will always have a higher score than the highest scored candidate generated from cand2. From the above pruning methods, only the third was not used by Z&C08.; It can be seen as an extra mechanism to help keep likely partial words in the agenda and improve the accuracy, but which does not give our system a speed advantage over Z&C08.; 3 Experiments We used the Chinese Treebank (CTB) data to perform one set of development tests and two sets of fi- Training iteration Figure 3: The influence of beam-sizes, and the convergence of the perceptron. nal tests. The CTB 4 was split into two parts, with the CTB 3 being used for a 10-fold cross validation test to compare speed and accuracies with Z&C08;, and the rest being used for development. The CTB 5 was used to perform the additional set of experiments to compare accuracies with other recent work. We use the standard F-measure to evaluate output accuracies. For word segmentation, precision is defined as the number of correctly segmented words divided by the total number of words in the output, and recall is defined as the number of correctly segmented words divided by the total number of words in the gold-standard output. For joint segmentation and POS-tagging, precision is defined as the number of correctly segmented and POS-tagged words divided by the total number of words from the output, and recall is defined as the correctly segmented and POS-tagged words divided by the total number of words in the gold-standard output. All our experiments were performed on a Linux platform, and a single 2.66GHz Intel Core 2 CPU. 3.1 Development tests Our development data consists of 150K words in 4798 sentences. 80% of the data were randomly chosen as the development training data, while the rest were used as the development test data. Our development tests were mainly used to decide the size ofthe beam, the number oftraining iterations, the ef- fect of partial features in beam-search decoding, and the effect of incremental learning (i.e. early update). 849 Figure 3 shows the accuracy curves for joint segmentation and POS-tagging by the number of training iterations, using different beam sizes. With the size of the beam increasing from 1to 32, the accuracies generally increase, while the amount of increase becomes small when the size of the beam becomes 16. After the 10th iteration, a beam size of 32 does not always give better accuracies than a beam size of 16. We therefore chose 16 as the size of the beam for our system. The testing times for each beam size between 1 and 32 are 7.16s, 11.90s, 18.42s, 27.82s, 46.77s and 89.21s, respectively. The corresponding speeds in the number of sentences per second are 111.45, 67.06, 43.32, 28.68, 17.06 and 8.95, respectively. Figure 3 also shows that the accuracy increases with an increased number of training iterations, but the amount of increase becomes small after the 25th iteration. We chose 29 as the number of iterations to train our system. The effect of incremental training: We compare the accuracies by incremental training using early update and normal perceptron training. In the normal perceptron training case, lines 11to 16 are taken out of the training algorithm in Figure 2. The algorithm reached the best performance at the 22nd iteration, with the segmentation F-score being 90.58% and joint F-score being 83.38%. In the incremental training case, the algorithm reached the best accuracy at the 30th training iteration, obtaining a segmentation F-score of 91.14% and a joint F-score of 84.06%. 3.2 Final tests using CTB 3 CTB 3 consists of 150K words in 10364 sentences. We follow Z&C08; and split it into 10 equal-sized parts. In each test, one part is taken as the test data and the other nine are combined together as the training data. We compare the speed and accuracy with the joint segmentor and tagger of Z&C08;, which is publicly available as the ZPar system, version 0.23. The results are shown in Table 2, where each row shows one cross validation test. The column head- ings “sf”, “jf”, “time” and “speed” refer to segmentation F-measure, joint F-measure, testing time (in 3http://www.sourceforge.net/projects/zpar #sZf&C08jftimespeed; tshfis papjefrtimespeed seconds) and testing speed (in the number of sentences per second), respectively. Our system gave a joint segmentation and POStagging F-score of 91.37%, which is only 0.04% lower than that of ZPar 0.2. The speed of our system was over 10 times as fast as ZPar 0.2. 3.3 Final tests using CTB 5 We follow Kruengkrai et al. (2009) and split the CTB 5 into training, development testing and testing sets, as shown in Table 3. We ignored the development test data since our system had been developed in previous experiments. Kruengkrai et al. (2009) made use of character type knowledge for spaces, numerals, symbols, alphabets, Chinese and other characters. In the previous experiments, our system did not use any knowledge beyond the training data. To make the comparison fairer, we included knowledge of English letters and Arabic numbers in this experiment. During both training and decoding, English letters and Arabic numbers are segmented using simple rules, treating consecutive English letters or Arabic numbers as a single word. The results are shown in Table 4, where row “N07” refers to the model of Nakagawa and Uchimoto (2007), rows “J08a” and “b” refer to the models of Jiang et al. (2008a) and Jiang et al. (2008b), and row “K09” refers to the models of Kruengkrai et al. (2009). Columns “sf” and “jf” refer to segmentation and joint accuracies, respectively. Our system 850 SectionsSentencesWords T Daerbsvltien3:gTrain14230i–7n021 g–71,3d90–21e035v 1elopm1e385n40t,8and5tes a648t,903o2n,18C92TB5. TJoKNab0ul8re79abs4(y:rtAesomcl-indurea)vycom9 pa7 r.i87s34o59n w3 i.t64h2710recntsudio sfjf CTB 5. gave comparable accuracies to these recent works, obtaining the best (same as the error-driven version of K09) joint F-score. 4 Related Work The effectiveness of our beam-search decoder showed that the joint segmentation and tagging problem may be less complex than previously perceived (Zhang and Clark, 2008; Jiang et al., 2008a). At the very least, the single model approach with a simple decoder achieved competitive accuracies to what has been achieved so far by the reranking (Shi and Wang, 2007; Jiang et al., 2008b) models and an ensemble model using machine-translation techniques (Jiang et al., 2008a). This may shed new light on joint segmentation and POS-tagging methods. Kruengkrai et al. (2009) and Zhang and Clark (2008) are the most similar to our system among related work. Both systems use a discriminatively trained linear model to score candidate outputs. The work of Kruengkrai et al. (2009) is based on Nakagawa and Uchimoto (2007), which separates the processing of known words and unknown words, and uses a set of segmentation tags to represent the segmentation of characters. In contrast, our model is conceptually simpler, and does not differentiate known words and unknown words. Moreover, our model is based on our previous work, in line with Zhang and Clark (2007), which does not treat word segmentation as character sequence labeling. Our learning and decoding algorithms are also different from Kruengkrai et al. (2009). While Kruengkrai et al. (2009) perform dynamic programming and MIRA learning, we use beam-search to perform incremental decoding, and the early-update version of the perceptron algorithm to train the model. Dynamic programming is exact inference, for which the time complexity is decided by the locality of feature templates. In contrast, beam-search is approximate and can run in linear time. The parameter updating for our algorithm is conceptually and computationally simpler than MIRA, though its performance can be slightly lower. However, the earlyupdate mechanism we use is consistent with our incremental approach, and improves the learning of the beam-search process. 5 Conclusion We showed that a simple beam-search decoding algorithm can be effectively applied to the decoding problem for a global linear model for joint word segmentation and POS-tagging. By guiding search with partial word information and performing learning for partial candidates, our system achieved sig- nificantly faster speed with little accuracy loss compared to the system of Z&C08.; The source code of our joint segmentor and POStagger can be found at: www.sourceforge.net/projects/zpar, version 0.4. 851 Acknowledgements We thank Canasai Kruengkrai for discussion on efficiency issues, and the anonymous reviewers for their suggestions. Yue Zhang and Stephen Clark are supported by the European Union Seventh Framework Programme (FP7-ICT-2009-4) under grant agreement no. 247762. References Eugene Charniak, Mark Johnson, Micha Elsner, Joseph Austerweil, David Ellis, Isaac Haxton, Catherine Hill, R. Shrivaths, Jeremy Moore, Michael Pozar, and Theresa Vu. 2006. Multilevel coarse-to-fine PCFG parsing. In Proceedings of HLT/NAACL, pages 168– 175, New York City, USA, June. Association for Computational Linguistics. Michael Collins and Brian Roark. 2004. Incremental parsing with the perceptron algorithm. In Proceedings of ACL, pages 111–1 18, Barcelona, Spain, July. Michael Collins. 2002. Discriminative training methods for hidden Markov models: Theory and experiments with perceptron algorithms. In Proceedings of EMNLP, pages 1–8, Philadelphia, USA, July. Wenbin Jiang, Liang Huang, Qun Liu, and Yajuan L u¨. 2008a. A cascaded linear model for joint Chinese word segmentation and part-of-speech tagging. In Proceedings of ACL/HLT, pages 897–904, Columbus, Ohio, June. Wenbin Jiang, Haitao Mi, and Qun Liu. 2008b. Word lattice reranking for Chinese word segmentation and part-of-speech tagging. In Proceedings of COLING, pages 385–392, Manchester, UK, August. Canasai Kruengkrai, Kiyotaka Uchimoto, Jun’ichi Kazama, Yiou Wang, Kentaro Torisawa, and Hitoshi Isahara. 2009. An error-driven word-character hybrid model for joint Chinese word segmentation and POS tagging. In Proceedings of ACL/AFNLP, pages 5 13– 521, Suntec, Singapore, August. Tetsuji Nakagawa and Kiyotaka Uchimoto. 2007. A hybrid approach to word segmentation and POS tagging. In Proceedings of ACL Demo and Poster Session, Prague, Czech Republic, June. Hwee Tou Ng and Jin Kiat Low. 2004. Chinese part-ofspeech tagging: One-at-a-time or all-at-once? word- based or character-based? In Proceedings of EMNLP, Barcelona, Spain. Brian Roark and Kristy Hollingshead. 2008. Classifying chart cells for quadratic complexity context-free inference. In Proceedings of COLING, pages 745– 752, Manchester, UK, August. Coling 2008 Organizing Committee. Yanxin Shi and Mengqiu Wang. 2007. A dual-layer CRF based joint decoding method for cascade segmentation and labelling tasks. In Proceedings of IJCAI, Hyderabad, India. Fei Xia, 2000. The part-of-speech tagging guidelines for the Chinese Treebank (3.0). Yue Zhang and Stephen Clark. 2007. Chinese segmentation with a word-based perceptron algorithm. In Proceedings of ACL, pages 840–847, Prague, Czech Republic, June. Yue Zhang and Stephen Clark. 2008. Joint word segmentation and POS tagging using a single perceptron. In Proceedings of ACL/HLT, pages 888–896, Columbus, Ohio, June. 852

4 0.13115655 25 emnlp-2010-Better Punctuation Prediction with Dynamic Conditional Random Fields

Author: Wei Lu ; Hwee Tou Ng

Abstract: This paper focuses on the task of inserting punctuation symbols into transcribed conversational speech texts, without relying on prosodic cues. We investigate limitations associated with previous methods, and propose a novel approach based on dynamic conditional random fields. Different from previous work, our proposed approach is designed to jointly perform both sentence boundary and sentence type prediction, and punctuation prediction on speech utterances. We performed evaluations on a transcribed conversational speech domain consisting of both English and Chinese texts. Empirical results show that our method outperforms an approach based on linear-chain conditional random fields and other previous approaches.

5 0.12972561 17 emnlp-2010-An Efficient Algorithm for Unsupervised Word Segmentation with Branching Entropy and MDL

Author: Valentin Zhikov ; Hiroya Takamura ; Manabu Okumura

Abstract: This paper proposes a fast and simple unsupervised word segmentation algorithm that utilizes the local predictability of adjacent character sequences, while searching for a leasteffort representation of the data. The model uses branching entropy as a means of constraining the hypothesis space, in order to efficiently obtain a solution that minimizes the length of a two-part MDL code. An evaluation with corpora in Japanese, Thai, English, and the ”CHILDES” corpus for research in language development reveals that the algorithm achieves an accuracy, comparable to that of the state-of-the-art methods in unsupervised word segmentation, in a significantly reduced . computational time.

6 0.097021297 41 emnlp-2010-Efficient Graph-Based Semi-Supervised Learning of Structured Tagging Models

7 0.073713444 33 emnlp-2010-Cross Language Text Classification by Model Translation and Semi-Supervised Learning

8 0.070207246 99 emnlp-2010-Statistical Machine Translation with a Factorized Grammar

9 0.069149904 30 emnlp-2010-Confidence in Structured-Prediction Using Confidence-Weighted Models

10 0.064834498 75 emnlp-2010-Lessons Learned in Part-of-Speech Tagging of Conversational Speech

11 0.064306892 10 emnlp-2010-A Probabilistic Morphological Analyzer for Syriac

12 0.063734025 119 emnlp-2010-We're Not in Kansas Anymore: Detecting Domain Changes in Streams

13 0.06009772 64 emnlp-2010-Incorporating Content Structure into Text Analysis Applications

14 0.056505535 48 emnlp-2010-Exploiting Conversation Structure in Unsupervised Topic Segmentation for Emails

15 0.053825676 39 emnlp-2010-EMNLP 044

16 0.053222552 62 emnlp-2010-Improving Mention Detection Robustness to Noisy Input

17 0.051045805 9 emnlp-2010-A New Approach to Lexical Disambiguation of Arabic Text

18 0.049821362 112 emnlp-2010-Unsupervised Discovery of Negative Categories in Lexicon Bootstrapping

19 0.049052805 79 emnlp-2010-Mining Name Translations from Entity Graph Mapping

20 0.047432486 13 emnlp-2010-A Simple Domain-Independent Probabilistic Approach to Generation

similar papers computed by lsi model

lsi for this paper:

topicId topicWeight

[(0, 0.191), (1, 0.091), (2, 0.009), (3, -0.043), (4, -0.366), (5, 0.023), (6, 0.077), (7, -0.086), (8, -0.221), (9, 0.018), (10, -0.049), (11, -0.076), (12, -0.027), (13, -0.131), (14, -0.111), (15, 0.117), (16, 0.175), (17, 0.017), (18, 0.08), (19, -0.13), (20, 0.179), (21, -0.165), (22, -0.101), (23, -0.006), (24, 0.05), (25, 0.13), (26, -0.007), (27, -0.148), (28, 0.017), (29, 0.044), (30, 0.04), (31, -0.027), (32, -0.04), (33, 0.077), (34, -0.036), (35, 0.065), (36, -0.036), (37, -0.015), (38, -0.083), (39, -0.031), (40, 0.039), (41, 0.031), (42, -0.044), (43, 0.006), (44, -0.008), (45, 0.046), (46, 0.064), (47, -0.041), (48, -0.01), (49, 0.002)]

similar papers list:

simIndex simValue paperId paperTitle

same-paper 1 0.96576494 43 emnlp-2010-Enhancing Domain Portability of Chinese Segmentation Model Using Chi-Square Statistics and Bootstrapping

Author: Baobao Chang ; Dongxu Han

2 0.77246314 2 emnlp-2010-A Fast Decoder for Joint Word Segmentation and POS-Tagging Using a Single Discriminative Model

Author: Yue Zhang ; Stephen Clark

3 0.75490278 69 emnlp-2010-Joint Training and Decoding Using Virtual Nodes for Cascaded Segmentation and Tagging Tasks

Author: Xian Qian ; Qi Zhang ; Yaqian Zhou ; Xuanjing Huang ; Lide Wu

4 0.61379814 17 emnlp-2010-An Efficient Algorithm for Unsupervised Word Segmentation with Branching Entropy and MDL

Author: Valentin Zhikov ; Hiroya Takamura ; Manabu Okumura

5 0.43328509 25 emnlp-2010-Better Punctuation Prediction with Dynamic Conditional Random Fields

Author: Wei Lu ; Hwee Tou Ng

6 0.32051712 41 emnlp-2010-Efficient Graph-Based Semi-Supervised Learning of Structured Tagging Models

7 0.31702483 10 emnlp-2010-A Probabilistic Morphological Analyzer for Syriac

8 0.27937689 75 emnlp-2010-Lessons Learned in Part-of-Speech Tagging of Conversational Speech

9 0.27752224 13 emnlp-2010-A Simple Domain-Independent Probabilistic Approach to Generation

10 0.25719404 30 emnlp-2010-Confidence in Structured-Prediction Using Confidence-Weighted Models

11 0.24832276 99 emnlp-2010-Statistical Machine Translation with a Factorized Grammar

12 0.24665566 119 emnlp-2010-We're Not in Kansas Anymore: Detecting Domain Changes in Streams

13 0.23930784 62 emnlp-2010-Improving Mention Detection Robustness to Noisy Input

14 0.23856352 45 emnlp-2010-Evaluating Models of Latent Document Semantics in the Presence of OCR Errors

15 0.22103325 9 emnlp-2010-A New Approach to Lexical Disambiguation of Arabic Text

16 0.21626803 48 emnlp-2010-Exploiting Conversation Structure in Unsupervised Topic Segmentation for Emails

17 0.20934974 108 emnlp-2010-Training Continuous Space Language Models: Some Practical Issues

18 0.20229706 33 emnlp-2010-Cross Language Text Classification by Model Translation and Semi-Supervised Learning

19 0.1857421 64 emnlp-2010-Incorporating Content Structure into Text Analysis Applications

20 0.18266444 79 emnlp-2010-Mining Name Translations from Entity Graph Mapping

similar papers computed by lda model

lda for this paper:

topicId topicWeight

[(1, 0.232), (3, 0.012), (12, 0.029), (29, 0.094), (30, 0.034), (32, 0.027), (52, 0.03), (56, 0.095), (62, 0.012), (64, 0.011), (66, 0.149), (72, 0.116), (76, 0.028), (77, 0.011), (79, 0.012), (87, 0.018), (89, 0.015)]

similar papers list:

simIndex simValue paperId paperTitle

same-paper 1 0.76028794 43 emnlp-2010-Enhancing Domain Portability of Chinese Segmentation Model Using Chi-Square Statistics and Bootstrapping

Author: Baobao Chang ; Dongxu Han

2 0.67138088 11 emnlp-2010-A Semi-Supervised Approach to Improve Classification of Infrequent Discourse Relations Using Feature Vector Extension

Author: Hugo Hernault ; Danushka Bollegala ; Mitsuru Ishizuka

Abstract: Several recent discourse parsers have employed fully-supervised machine learning approaches. These methods require human annotators to beforehand create an extensive training corpus, which is a time-consuming and costly process. On the other hand, unlabeled data is abundant and cheap to collect. In this paper, we propose a novel semi-supervised method for discourse relation classification based on the analysis of cooccurring features in unlabeled data, which is then taken into account for extending the feature vectors given to a classifier. Our experimental results on the RST Discourse Treebank corpus and Penn Discourse Treebank indicate that the proposed method brings a significant improvement in classification accuracy and macro-average F-score when small training datasets are used. For instance, with training sets of c.a. 1000 labeled instances, the proposed method brings improvements in accuracy and macro-average F-score up to 50% compared to a baseline classifier. We believe that the proposed method is a first step towards detecting low-occurrence relations, which is useful for domains with a lack of annotated data.

3 0.66836208 69 emnlp-2010-Joint Training and Decoding Using Virtual Nodes for Cascaded Segmentation and Tagging Tasks

Author: Xian Qian ; Qi Zhang ; Yaqian Zhou ; Xuanjing Huang ; Lide Wu

4 0.66511917 49 emnlp-2010-Extracting Opinion Targets in a Single and Cross-Domain Setting with Conditional Random Fields

Author: Niklas Jakob ; Iryna Gurevych

Abstract: In this paper, we focus on the opinion target extraction as part of the opinion mining task. We model the problem as an information extraction task, which we address based on Conditional Random Fields (CRF). As a baseline we employ the supervised algorithm by Zhuang et al. (2006), which represents the state-of-the-art on the employed data. We evaluate the algorithms comprehensively on datasets from four different domains annotated with individual opinion target instances on a sentence level. Furthermore, we investigate the performance of our CRF-based approach and the baseline in a single- and cross-domain opinion target extraction setting. Our CRF-based approach improves the performance by 0.077, 0.126, 0.071 and 0. 178 regarding F-Measure in the single-domain extraction in the four domains. In the crossdomain setting our approach improves the performance by 0.409, 0.242, 0.294 and 0.343 regarding F-Measure over the baseline.

5 0.65761149 35 emnlp-2010-Discriminative Sample Selection for Statistical Machine Translation

Author: Sankaranarayanan Ananthakrishnan ; Rohit Prasad ; David Stallard ; Prem Natarajan

Abstract: Production of parallel training corpora for the development of statistical machine translation (SMT) systems for resource-poor languages usually requires extensive manual effort. Active sample selection aims to reduce the labor, time, and expense incurred in producing such resources, attaining a given performance benchmark with the smallest possible training corpus by choosing informative, nonredundant source sentences from an available candidate pool for manual translation. We present a novel, discriminative sample selection strategy that preferentially selects batches of candidate sentences with constructs that lead to erroneous translations on a held-out development set. The proposed strategy supports a built-in diversity mechanism that reduces redundancy in the selected batches. Simulation experiments on English-to-Pashto and Spanish-to-English translation tasks demon- strate the superiority of the proposed approach to a number of competing techniques, such as random selection, dissimilarity-based selection, as well as a recently proposed semisupervised active learning strategy.

6 0.65627217 82 emnlp-2010-Multi-Document Summarization Using A* Search and Discriminative Learning

7 0.65601939 45 emnlp-2010-Evaluating Models of Latent Document Semantics in the Presence of OCR Errors

8 0.65330166 65 emnlp-2010-Inducing Probabilistic CCG Grammars from Logical Form with Higher-Order Unification

9 0.6524927 120 emnlp-2010-What's with the Attitude? Identifying Sentences with Attitude in Online Discussions

10 0.65070194 78 emnlp-2010-Minimum Error Rate Training by Sampling the Translation Lattice

11 0.64913523 58 emnlp-2010-Holistic Sentiment Analysis Across Languages: Multilingual Supervised Latent Dirichlet Allocation

12 0.64888155 117 emnlp-2010-Using Unknown Word Techniques to Learn Known Words

13 0.64745885 2 emnlp-2010-A Fast Decoder for Joint Word Segmentation and POS-Tagging Using a Single Discriminative Model

14 0.6472162 32 emnlp-2010-Context Comparison of Bursty Events in Web Search and Online Media

15 0.6462177 60 emnlp-2010-Improved Fully Unsupervised Parsing with Zoomed Learning

16 0.64619797 25 emnlp-2010-Better Punctuation Prediction with Dynamic Conditional Random Fields

17 0.64450622 86 emnlp-2010-Non-Isomorphic Forest Pair Translation

18 0.64281309 103 emnlp-2010-Tense Sense Disambiguation: A New Syntactic Polysemy Task

19 0.64271367 84 emnlp-2010-NLP on Spoken Documents Without ASR

20 0.64128876 20 emnlp-2010-Automatic Detection and Classification of Social Events