During crawling and indexing, Fess:

1. Detects language from document content using Apache Tika
2. Validates against `supported.languages` list
3. Creates language-specific fields (e.g., `content_ja`, `title_en`, `content_sv`)
4. Applies language-specific analyzers for better search results

Configured in `fess_config.properties`:
```properties
indexer.language.fields=content,important_content,title
```

        // Test with Japanese characters
        String word = "日本語テスト";
        float boost = 2.0f;
        List<String> readings = Arrays.asList("ニホンゴテスト");
        List<String> fields = Arrays.asList("content_ja");
        List<String> tags = Arrays.asList("japanese");
        List<String> roles = Arrays.asList("admin");

        elevateWord = new ElevateWord(word, boost, readings, fields, tags, roles);