@techreport{oai:ipsj.ixsq.nii.ac.jp:00081867,
 author = {辻, 理絵子 and 木村, 健 and 古宮, 嘉那子 and 小谷, 善行 and Rieko, Tsuji and Takeshi, Kimura and Kanako, Komiya and Yoshiyuki, Kotani},
 issue = {2},
 month = {May},
 note = {外国人の商外国人が日本語のショッピングサイトを利用する際日本語の商品名が分からず，商品が見つけられないことがある．本研究では，実際に外国人が日本のショッピングサイトで商品検索の際に失敗したクエリとその正しいクエリの対のコーパスを用いて，非日本語圏のユーザによる検索クエリの音訳を行った．ペアコーパスには，意訳によって修正されたクエリのように，音訳するにはノイズとなるデータが含まれているため，文字種によるフィルタリングを行って学習データを絞り込んだ．さらに，BIGRAM，HMM，CRF の 3 つの機械翻訳手法を比較した結果，検索クエリの音訳では HMM 手法が最適であった．, There are some cases where the non-Japanese buyers are unable to find products they want through the Japanese shopping Web sites because it requires Japanese queries. We propose to transliterate the inputs of the non-Japanese user, i.e., search queries written in English alphabets, into Japanese Katakana to solve that problem. In this research, the pairs of the non-Japanese search queries which failed to get the right match obtained from a Japanese shopping website and its transcribed word given by volunteers were used for the training data. Since this corpus includes some noises for transliteration such as free translation, we used two different filters to filter out the query pairs that are not transliterated in order to improve the quality of the training data. In addition, we compared three methods, i.e., BIGRAM, HMM, and CRF, using these data to investigate which transliteration method is the best for query transliteration. The experiment revealed the HMM was the best.},
 title = {外国人の検索クエリに対する音訳手法の適用},
 year = {2012}
}