Suggesting Sentences for ESL using Kernel Embeddings

Kent Shioda, Mamoru Komachi, Rue Ikeya, Daichi Mochihashi


Abstract
Sentence retrieval is an important NLP application for English as a Second Language (ESL) learners. ESL learners are familiar with web search engines, but generic web search results may not be adequate for composing documents in a specific domain. However, if we build our own search system specialized to a domain, it may be subject to the data sparseness problem. Recently proposed word2vec partially addresses the data sparseness problem, but fails to extract sentences relevant to queries owing to the modeling of the latent intent of the query. Thus, we propose a method of retrieving example sentences using kernel embeddings and N-gram windows. This method implicitly models latent intent of query and sentences, and alleviates the problem of noisy alignment. Our results show that our method achieved higher precision in sentence retrieval for ESL in the domain of a university press release corpus, as compared to a previous unsupervised method used for a semantic textual similarity task.
Anthology ID:
W17-5911
Volume:
Proceedings of the 4th Workshop on Natural Language Processing Techniques for Educational Applications (NLPTEA 2017)
Month:
December
Year:
2017
Address:
Taipei, Taiwan
Venues:
NLP-TEA | WS
SIG:
Publisher:
Asian Federation of Natural Language Processing
Note:
Pages:
64–68
Language:
URL:
https://aclanthology.org/W17-5911
DOI:
Bibkey:
Cite (ACL):
Kent Shioda, Mamoru Komachi, Rue Ikeya, and Daichi Mochihashi. 2017. Suggesting Sentences for ESL using Kernel Embeddings. In Proceedings of the 4th Workshop on Natural Language Processing Techniques for Educational Applications (NLPTEA 2017), pages 64–68, Taipei, Taiwan. Asian Federation of Natural Language Processing.
Cite (Informal):
Suggesting Sentences for ESL using Kernel Embeddings (Shioda et al., 2017)
Copy Citation:
PDF:
https://aclanthology.org/W17-5911.pdf