Pivot Approach for Extracting Paraphrase Patterns from Bilingual Corpora
Zhao, Shiqi and Wang, Haifeng and Liu, Ting and Li, Sheng

Article Structure

Abstract

Paraphrase patterns are useful in paraphrase recognition and generation.

Introduction

Paraphrases are different expressions that convey the same meaning.

Related Work

Paraphrase patterns have been learned and used in information extraction (IE) and answer extraction of QA.

Proposed Method

3.1 Corpus Preprocessing

Experiments

The EC parallel corpus in our experiments was constructed using several LDC bilingual corpora5.

Conclusion

This paper proposes a pivot approach for extracting paraphrase patterns from bilingual corpora.

Topics

log-linear

Appears in 12 sentences as: Log-linear (1) log-linear (11)
In Pivot Approach for Extracting Paraphrase Patterns from Bilingual Corpora
  1. We propose a log-linear model to compute the paraphrase likelihood of two patterns and exploit feature functions based on maximum likelihood estimation (MLE) and lexical weighting (LW).
    Page 1, “Abstract”
  2. parsing and English-foreign language word alignment, (2) aligned patterns induction, which produces English patterns along with the aligned pivot patterns in the foreign language, (3) paraphrase patterns extraction, in which paraphrase patterns are extracted based on a log-linear model.
    Page 2, “Introduction”
  3. Secondly, we propose a log-linear model for computing the paraphrase likelihood.
    Page 2, “Introduction”
  4. Besides, the log-linear model is more effective than the conventional model presented in (Bannard and Callison-Burch, 2005).
    Page 2, “Introduction”
  5. In order to exploit more and richer information to estimate the paraphrase likelihood, we propose a log-linear model:
    Page 4, “Proposed Method”
  6. In this paper, 4 feature functions are used in our log-linear model, which include:
    Page 5, “Proposed Method”
  7. 4.1 Evaluation of the Log-linear Model
    Page 6, “Experiments”
  8. As previously mentioned, in the log-linear model of this paper, we use both MLE based and LW based feature functions.
    Page 6, “Experiments”
  9. In this section, we evaluate the log-linear model (LL-Model) and compare it with the MLE based model (MLE-Model) presented by Bannard and Callison-Burch (2005)6.
    Page 6, “Experiments”
  10. We use a log-linear model to compute the paraphrase likelihood and exploit feature functions based on MLE and LW.
    Page 8, “Conclusion”
  11. In addition, the log-linear model with the proposed feature functions significantly outperforms the conventional models.
    Page 8, “Conclusion”

See all papers in Proc. ACL 2008 that mention log-linear.

See all papers in Proc. ACL that mention log-linear.

Back to top.

log-linear model

Appears in 12 sentences as: Log-linear Model (1) log-linear model (10) log-linear model: (1)
In Pivot Approach for Extracting Paraphrase Patterns from Bilingual Corpora
  1. We propose a log-linear model to compute the paraphrase likelihood of two patterns and exploit feature functions based on maximum likelihood estimation (MLE) and lexical weighting (LW).
    Page 1, “Abstract”
  2. parsing and English-foreign language word alignment, (2) aligned patterns induction, which produces English patterns along with the aligned pivot patterns in the foreign language, (3) paraphrase patterns extraction, in which paraphrase patterns are extracted based on a log-linear model .
    Page 2, “Introduction”
  3. Secondly, we propose a log-linear model for computing the paraphrase likelihood.
    Page 2, “Introduction”
  4. Besides, the log-linear model is more effective than the conventional model presented in (Bannard and Callison-Burch, 2005).
    Page 2, “Introduction”
  5. In order to exploit more and richer information to estimate the paraphrase likelihood, we propose a log-linear model:
    Page 4, “Proposed Method”
  6. In this paper, 4 feature functions are used in our log-linear model , which include:
    Page 5, “Proposed Method”
  7. 4.1 Evaluation of the Log-linear Model
    Page 6, “Experiments”
  8. As previously mentioned, in the log-linear model of this paper, we use both MLE based and LW based feature functions.
    Page 6, “Experiments”
  9. In this section, we evaluate the log-linear model (LL-Model) and compare it with the MLE based model (MLE-Model) presented by Bannard and Callison-Burch (2005)6.
    Page 6, “Experiments”
  10. We use a log-linear model to compute the paraphrase likelihood and exploit feature functions based on MLE and LW.
    Page 8, “Conclusion”
  11. In addition, the log-linear model with the proposed feature functions significantly outperforms the conventional models.
    Page 8, “Conclusion”

See all papers in Proc. ACL 2008 that mention log-linear model.

See all papers in Proc. ACL that mention log-linear model.

Back to top.

word alignment

Appears in 5 sentences as: word alignment (5)
In Pivot Approach for Extracting Paraphrase Patterns from Bilingual Corpora
  1. parsing and English-foreign language word alignment , (2) aligned patterns induction, which produces English patterns along with the aligned pivot patterns in the foreign language, (3) paraphrase patterns extraction, in which paraphrase patterns are extracted based on a log-linear model.
    Page 2, “Introduction”
  2. We conduct word alignment with Giza++ (Och and Ney, 2000) in both directions and then apply the grow-diag heuristic (Koehn et al., 2005) for symmetrization.
    Page 3, “Proposed Method”
  3. where a denotes the word alignment between 6 and 6. n is the number of words in 6.
    Page 5, “Proposed Method”
  4. It is not surprising, since Bannard and Callison-Burch (2005) have pointed out that word alignment error is the major factor that influences the performance of the methods learning paraphrases from bilingual corpora.
    Page 6, “Experiments”
  5. The LW based features validate the quality of word alignment and assign low scores to those aligned EC pattern pairs with incorrect alignment.
    Page 6, “Experiments”

See all papers in Proc. ACL 2008 that mention word alignment.

See all papers in Proc. ACL that mention word alignment.

Back to top.

data sparseness

Appears in 4 sentences as: data sparseness (4)
In Pivot Approach for Extracting Paraphrase Patterns from Bilingual Corpora
  1. However, we find that using only the MLE based probabilities can suffer from data sparseness .
    Page 4, “Proposed Method”
  2. To alleviate the data sparseness problem, we only kept patterns appearing more than 10 times in the corpus for extracting paraphrase patterns.
    Page 6, “Experiments”
  3. In other words, it seriously suffers from data sparseness .
    Page 7, “Experiments”
  4. which is mainly because the data sparseness problem is more serious when extracting long patterns.
    Page 8, “Experiments”

See all papers in Proc. ACL 2008 that mention data sparseness.

See all papers in Proc. ACL that mention data sparseness.

Back to top.

LM

Appears in 4 sentences as: LM (5)
In Pivot Approach for Extracting Paraphrase Patterns from Bilingual Corpora
  1. In detail, a paraphrase pattern 6’ of e was reranked based on a language model ( LM ):
    Page 8, “Experiments”
  2. scoreLM(e’ |SE) is the LM based score: scoreLM(e’|SE) = %logPLM(S’E), where 8% is the sentence generated by replacing e in SE with e’ .
    Page 8, “Experiments”
  3. To investigate the contribution of the LM based score, we ran the experiment again with A = l (ignoring the LM based score) and found that the precision is 57.09%.
    Page 8, “Experiments”
  4. It indicates that the LM based reranking can improve the precision.
    Page 8, “Experiments”

See all papers in Proc. ACL 2008 that mention LM.

See all papers in Proc. ACL that mention LM.

Back to top.

sentence pairs

Appears in 4 sentences as: sentence pairs (4)
In Pivot Approach for Extracting Paraphrase Patterns from Bilingual Corpora
  1. Using the presented method, we extract over 1,000,000 pairs of paraphrase patterns from 2M bilingual sentence pairs , the precision of which exceeds 67%.
    Page 1, “Abstract”
  2. Using the proposed approach, we extract over 1,000,000 pairs of paraphrase patterns from 2M bilingual sentence pairs , the precision of which is above 67%.
    Page 2, “Introduction”
  3. Thus the method acquired paraphrase patterns from sentence pairs that share comparable NEs.
    Page 2, “Related Work”
  4. Experimental results show that the pivot approach is effective, which extracts over 1,000,000 pairs of paraphrase patterns from 2M bilingual sentence pairs .
    Page 8, “Conclusion”

See all papers in Proc. ACL 2008 that mention sentence pairs.

See all papers in Proc. ACL that mention sentence pairs.

Back to top.

subtrees

Appears in 4 sentences as: subtrees (5)
In Pivot Approach for Extracting Paraphrase Patterns from Bilingual Corpora
  1. To induce the aligned patterns, we first induce the English patterns using the subtrees and partial subtrees .
    Page 3, “Proposed Method”
  2. 1Note that, a subtree may contain several partial subtrees .
    Page 3, “Proposed Method”
  3. In this paper, all the possible partial subtrees are considered when extracting paraphrase patterns.
    Page 3, “Proposed Method”
  4. First, only subtrees containing no more than 10 words were used to induce English patterns.
    Page 5, “Experiments”

See all papers in Proc. ACL 2008 that mention subtrees.

See all papers in Proc. ACL that mention subtrees.

Back to top.

context information

Appears in 3 sentences as: context information (3)
In Pivot Approach for Extracting Paraphrase Patterns from Bilingual Corpora
  1. In Section 4.1, we have evaluated the precision of the paraphrase patterns without considering context information .
    Page 8, “Experiments”
  2. The context information was also considered by our judges.
    Page 8, “Experiments”
  3. In addition, we will try to make better use of the context information when replacing paraphrase patterns in context sentences.
    Page 8, “Conclusion”

See all papers in Proc. ACL 2008 that mention context information.

See all papers in Proc. ACL that mention context information.

Back to top.

parallel corpus

Appears in 3 sentences as: parallel corpus (3)
In Pivot Approach for Extracting Paraphrase Patterns from Bilingual Corpora
  1. An English-Chinese (EC) bilingual parallel corpus is employed for training.
    Page 3, “Proposed Method”
  2. The EC parallel corpus in our experiments was constructed using several LDC bilingual corpora5.
    Page 5, “Experiments”
  3. In our experiment, we implemented DIRT and extracted paraphrase patterns from the English part of our bilingual parallel corpus .
    Page 6, “Experiments”

See all papers in Proc. ACL 2008 that mention parallel corpus.

See all papers in Proc. ACL that mention parallel corpus.

Back to top.

parse tree

Appears in 3 sentences as: parse tree (2) parse trees (1)
In Pivot Approach for Extracting Paraphrase Patterns from Bilingual Corpora
  1. Let SE be an English sentence, TE the parse tree of SE, 6 a word of SE, we define the subtree and partial subtree following the definitions in (Ouan-graoua et al., 2007).
    Page 3, “Proposed Method”
  2. If e,-is a descendant of ej in the parse tree , we remove p05,- from PE(e).
    Page 4, “Proposed Method”
  3. Note that the Chinese patterns are not extracted from parse trees .
    Page 4, “Proposed Method”

See all papers in Proc. ACL 2008 that mention parse tree.

See all papers in Proc. ACL that mention parse tree.

Back to top.

significantly outperforms

Appears in 3 sentences as: significantly outperforms (3)
In Pivot Approach for Extracting Paraphrase Patterns from Bilingual Corpora
  1. The evaluation results show that: (l) The pivot approach is effective in extracting paraphrase patterns, which significantly outperforms the conventional method DIRT.
    Page 1, “Abstract”
  2. Our experiments show that the pivot approach significantly outperforms conventional methods.
    Page 2, “Introduction”
  3. In addition, the log-linear model with the proposed feature functions significantly outperforms the conventional models.
    Page 8, “Conclusion”

See all papers in Proc. ACL 2008 that mention significantly outperforms.

See all papers in Proc. ACL that mention significantly outperforms.

Back to top.

PDF available from www.aclweb.org.
Send us an email