テキストデ タを使った音声認識誤りの訂正  [in Japanese] Correction of Speech Recognition Error using Text Data  [in Japanese]

Access this Article

Search this Article

Author(s)

Abstract

認識誤りに起因して, 音声翻訳の性能 (品質) が劣化するという問題がある. 認識結果の正解部分のみを翻訳する手法が提案されているが, 翻訳されない部分に関する情報は失われてしまう. 我々はこの問題を解決するため, 次のような手順からなる誤り訂正の手法を提案する.(1) 訂正の必要性の判断および誤り部分の特定を行う.(2) 認識結果の誤り近傍に関して音韻的に近い用例をテキストデータ中から検索し, 訂正候補の生成を行う.(3) 訂正候補の妥当性を意味と音韻の両方の観点から判断し, 最も妥当なものを選択する. 提案手法を音声翻訳システムに組み込み, 旅行会話を対象として評価した. 認識結果の単語誤り率で2.3%の減少, 翻訳率で5.4%の増加が得られ, 提案手法の有効性が示された.

Because of recognition errors, the performance (quality) of speech translation is degraded. Previously, we proposed a method that used only reliable parts of the recognition result for the translation. However, in this method, non-translated parts are omitted even if useful information exist in these parts. To overcome this problem, we propose an error correction method which is composed of the following steps: (1) The necessity of correction is judged and only utterances of the recognition results with “potentially” recoverable erroneous parts are retained.(2) The example utterances that have phonetically similar parts to the ones retained in the step (1) are retrieved from a text corpus, and correction hypotheses are created.(3) The reliability of the correction hypotheses is judged according to both semantic and phonetic point of view and the most reliable one is selected. The error correction method was incorporated into a speech translation system, and evaluated for speech inputs in travel conversations. As the results, the word error rate was reduced by 2.3%, and the acceptable translations rate was increased by 5.4%.

Journal

  • Journal of Natural Language Processing

    Journal of Natural Language Processing 7(4), 205-227, 2000-10-10

    The Association for Natural Language Processing

References:  22

Cited by:  12

Codes

  • NII Article ID (NAID)
    10008830021
  • NII NACSIS-CAT ID (NCID)
    AN10472659
  • Text Lang
    JPN
  • Article Type
    Journal Article
  • ISSN
    13407619
  • NDL Article ID
    5544296
  • NDL Call No.
    Z21-B168
  • Data Source
    CJP  CJPref  NDL  J-STAGE 
Page Top