日本語 NINJAL
 

XML Document Structure

It is possible to look up a large amount of data regarding the texts collected in the BCCWJ, as illustrated below.

  • Bibliographical Information

    Citation information (Title, Author Name, Publishing Company, Publication Year, Genre)
  • Japanese character information

    Character readings, misprints, JIS encoding, calculation superscripting, subscript information
  • Structure Information

    Information regarding the layered structure of the text (article, paragraph, sentence, captions, quotations, itemized lists, etc.)
  • Data Sampling Information

    Information about the scope of Fixed-length samples.

Description of XML

The above information is compiled using XML markup. For example, take the following text:

sashie1

The above text will then be digitized as in the following example. The sections enclosed in < > are called tags. What type of tags are available and how they are organized has a large influence on how information can be accessed in the corpus.

sashie2

Tag Types and Meanings

This section will explain the tags used in the BCCWJ.

 Tag NameContents
Samples sample The scope of a single sample.
sampling Information related to the sampling points.
Layered Structure
(Document Structure)
article Identifies author and theme.
title Gives a title descriptive of the contents, such as a chapter or article title.
cluster Marks the whole text of a title tag.
list Itemized lists or lists of noun phrases. For list elements.
paragraph Marks the boundaries of paragraphs.
sentence Marks sentence boundaries.
Charts

(Document Structure)
figure Figures・Charts・Photos・Pictures, and others
caption Titles and explanations of figures.
Citations
(Document Structure)
citation Any citations of other documents.
speech Speech and internal monologues, opening sentences.
Annotation
(Document Structure)
noteBody Footnotes, endnotes, etc. An element describing or annotating the original text.
Other
(Document Structure)
abstract Outlines that do not fit under the article or cluster tags.
verse Verses of poems、waka、haiku、songs, etc.
Characters and Transcription ruby Readings of Kanji characters
correction Corrections of mistakes in the original text.
missingCharacter Characters not included in the standard encoding (non-JIS).

Character Encoding

Digitized texts are encoded using the JISX0213:2004 standard (aka JIS 4th edition), a type of Unicode encoding.