いわゆる 「異体字」 の概念を漢字の派生関係と通用関係とに整理した上で，人間文化研究機構研究資源共有化統合検索システムでの運用を想定し，検索のための必要最低限の 「異体字」 群を収録した異体漢字対応テーブルの作成事例を報告する。This paper discusses the concept of kanji variants in terms of relationships based on formal derivation, and customary usage, and reports on the production of an optimized table listing kanji variants for use with the Resource Sharing System for the Humanities, with excludes sets of kanji variants, particularly customary usage kanji variants, that can lead to unexpected search results.