ASPEC: Asian Scientific Paper Excerpt Corpus

By Toshiaki Nakazawa, Manabu Yaguchi, Kiyotaka Uchimoto, Masso Utiyama, Eiichiro Sumita, Sadao Kurohashi, Hitoshi Isahara. 2016

224 citations recorded by the source corpus.

In this paper, we describe the details of the ASPEC (Asian Scientific Paper Excerpt Corpus), which is the first large-size parallel corpus of scientific paper domain. ASPEC was constructed in the Japanese-Chinese machine translation project conducted between 2006 and 2010 using the Special Coordination Funds for Promoting Science and Technology. It consists of a Japanese-English scientific paper abstract corpus of approximately 3 million parallel sentences (ASPEC-JE) and a Chinese-Japanese scientific paper excerpt corpus of approximately 0.68 million parallel sentences (ASPEC-JC). ASPEC is used as the official dataset for the machine translation evaluation workshop WAT (Workshop on Asian Translation).

Subfields: Artificial Intelligence, Natural Language Processing, Evaluation & Benchmarks

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

In Evaluation & Benchmarks

Citations in the corpus

Structured Attention Networks (2017, neutral): “The data comes from the Workshop on Asian Translation (WAT) (Nakazawa et al., 2016).”
MURAL: Multimodal, Multitask Retrieval Across Languages (2021, background): “0, Autshumato (Groe-newald and du Plooy, 2010), PMIndia (Haddow and Kirefu, 2020), CVIT (Srivastava et al., 2020), Inuktitut (Hernandez and Nguyen, 2020), NLPC, JESC (Pryzant et al., 2018), KFTT (Neubig, 2011), ASPEC (Nakazawa et al., 2016), Flores (Guzmán et al., 2019).”
An Empirical Comparison of Domain Adaptation Methods for Neural Machine Translation (2017, methodology): “The scientific domain MT was conducted on the Chinese-Japanese paper excerpt corpus (ASPEC-CJ)3 (Nakazawa et al., 2016), which is one subtask of the workshop on Asian translation (WAT)4 (Nakazawa et al.”
An Empirical Comparison of Simple Domain Adaptation Methods for Neural Machine Translation (2017, background): “One of the most attractive features of neural machine translation (NMT) (Bahdanau et al., 2015; Cho et al., 2014; Sutskever et al., 2014) is that it is possible to train an end to end system without the need to deal with word alignments, translation rules and complicated decoding algorithms, which…”
Improved Neural Machine Translation with Chinese Phonologic Features (2018, methodology): “IntheJapanese-Chinese task, we use 2.87M sentence pairs from ASPEC Japanese-Chinese corpus [15] 2 .”
Incorporating Discrete Translation Lexicons into Neural Machine Translation (2016, methodology): “To test whether the proposed method is useful on larger data sets, we also performed follow-up experiments on the larger Japanese-English ASPEC dataset (Nakazawa et al., 2016) that consist of 2 million training examples, 63 million tokens, and 81,000 vocabulary size.”
Controlling the Voice of a Sentence in Japanese-to-English Neural Machine Translation (2016, background): “Table 4 shows the results of two methods submitted for the shared task at WAT 2016 (Nakazawa et al., 2016a).”
Chunk-based Decoder for Neural Machine Translation (2017, methodology): “Proposed Models vs. Baselines Table 3 shows the experimental results on the ASPEC test set.”
Data augmentation using back-translation for context-aware neural machine translation (2019, background): “…the last possibility ( § 6), since there are few parallel corpora for context-aware translation; existing large-scale and high-quality parallel corpora are usually obtained by extracting reliable sentence alignments from translations by humans (Nakazawa et al., 2016; Pryzant et al., 2018).”
On the Relation between Position Information and Sentence Length in Neural Machine Translation (2019, methodology): “For English-to-Japanese translation task, we exploit ASPEC (Nakazawa et al., 2016), a parallel corpus compiled from abstract sections of scientiﬁc papers.”
Neural Machine Translation via Binary Code Prediction (2017, unknown): “In addition, we can see that the BLEU score in BTEC quickly improves, and saturates at N = 1024 in contrast to the ASPEC model, which is still improving at N = 2048 .”
Lexicons and Minimum Risk Training for Neural Machine Translation: NAIST-CMU at WAT2016 (2016, methodology): “The system was evaluated according to the official WAT “HUMAN” metric (Nakazawa et al., 2016a), which consistsof pairwise comparisons with a baseline phrase-based system, where the evaluated system receives +1 for every win, -1 for every tie, 0 for every loss, these values are averaged over all…”
Sequence-to-Dependency Neural Machine Translation (2017, methodology): “The experiments are conducted on the ChineseEnglish task as well as the Japanese-English translation tasks where the same data set from WAT 2016 ASPEC corpus (Nakazawa et al., 2016) 3 is used for a fair comparison with other work.”
Reward Optimization for Neural Machine Translation with Learned Metrics (2021, methodology): “Different from our approach, Zhu et al. (2020) WMT19 Ru-En (Barrault et al., 2019) and ASPEC Ja-En (Nakazawa et al., 2016).”
Multilingual Extractive Reading Comprehension by Runtime Machine Translation (2018, methodology): “However, in our preliminary experiments, we found that even a state-of-the-art Japanese-to-English NMT model trained with the ASPEC corpus (Nakazawa et al. 2016), an established English-toJapanese scientific corpus, resulting in poor performance in translating Wikipedia articles.”
An Empirical Study of Adequate Vision Span for Attention-Based Neural Machine Translation (2016, methodology): “The training data of En-Ja task is based on ASPEC parallel corpus (Nakazawa et al., 2016), which contains 3M sentence pairs, whereas the test data contains 1812 sentences, which have 24.4 words on average.”
An Empirical Study of Mini-Batch Creation Strategies for Neural Machine Translation (2017, methodology): “We carried out experiments with two language pairs, English-Japanese using the ASPEC-JE cor- 2https://github.com/rsennrich/nematus 3https://github.com/neubig/lamtram pus (Nakazawa et al., 2016) and English-German using the WMT 2016 news task with newstest2016 as the test-set (Bojar et al., 2016).”
JESC: Japanese-English Subtitle Corpus (2017, methodology): “The JPO is a 5- point system which provides strong guidelines for scoring the quality of a Japanese-English translation pair (Nakazawa et al., 2016).”
Neural Machine Translation with Source-Side Latent Graph Parsing (2017, methodology): “We used an English-to-Japanese translation task of the Asian Scientiﬁc Paper Excerpt Corpus (AS-PEC) (Nakazawa et al., 2016b) used in the Work-shop on Asian Translation (WAT), since it has been shown that syntactic information is useful in English-to-Japanese translation (Eriguchi et al., 2016b; Neu”