We have studied the automatic construction of a multilingual citation index by collecting Postscript and PDF files from the Internet. We propose a method to identify duplicate bibliographic information written in both Japanese and English, which will be an indispensable module for the construction of a multilingual citation index. To confirm the effectiveness of our method, we conducted an examination and found that our method obtained a precision of 94.8% and a recall of 25.6%. Key words: translation of technical terms, multilingual citation index, identification of interlingual bibliographic information.