Technical platform & interface

The digitized Shen bao newspaper is available on two different platforms, either Green Apple or Huawenku, which is produced by Naver on behalf of Green Apple. Each of the two platforms offer their own strengths and weaknesses. It would be ideal in the future to combine the better features of the two platforms. One major interface challenge noted during trial access was the inability to download or print page images, which would be a considerable hindrance.  But East View notes that downloading of both transcribed text files and page images will be available in the actual database.

The following platform characteristics have been observed by librarians at CRL institutions during trial access:

Green Apple strengths:

• The interface design is clean, and searching by both dates and articles seems to be quick, straightforward and intuitive.
• The search interface has more options than Huawenku (by date, main text, title/ headline, author, column, etc.), and the options are displayed all at once in full view (not hidden as is the case with Huawenku).
• A search in the main text by a keyword or phrase provides precise (not fuzzy) results.

Green Apple weaknesses:

• There is only minimal English language labeling in the interface.
• Navigation between the various functions of the site is awkward; there are few links back to the home page or the main search and browse pages.
• Searches are allowed only in traditional (not simplified) Chinese.
• The date browse function is not very flexible. The calendar option for browsing dates could be more prominently displayed, since otherwise one must browse through a long list of dates.
• It would help if the page images had better text legibility; perhaps some clarity has been lost due to the high contrast of the images, but some may be due to the quality of the original printing or the microfilm source. The images themselves have fairly high pixel count (averaging nearly 7 megabytes.)
• Pages are not divided (zoned) into different sections by article component; instead zoomable full page images are presented using the open source document viewer "Flexpaper, similar to Adobe Acrobat.
• The displayed plain text reveals errors both in main texts (poor OCR?) and author names (mistyped transcriptions?).

Huawenku Platform

Huawenku strengths:

• Page images are divided (zoned) according to different sections such as news or articles, and can be individually downloaded and printed.
• Searches in both traditional and simplified Chinese are allowed, and users can switch between the two styles in the displayed results text.
• Users can use and input hand-written and matched characters in the search box.
• There are several easy to use options for date browsing, including a visual timeline, and a very flexible calendar view.

Huawenku weaknesses:

• The interface is in Chinese language only.
• The legibility of the page images appears essentially the same as in the Green Apple interface, but they may have a more crisp appearance, due to the image viewer used.
• Its search interface provides fewer options than the Green Apple interface, and the options are hidden and sometimes confusing to users.
• Search (or fuzzy search) in “All” by a keyword or phrase yields a very large number of results; it does not seem to provide a precision search.
• Search by date requires input of words.
• The displayed plain text does not provide the author’s name for articles or news reports (even when it is available on the original page).
• There seem to be a number of cases where items appearing in the page images of the original newspaper are not zoned and searchable in the database. For instance, on page 15, November 16, 1933, there are at least 6 articles, but the database show “full-text” as “soft text” (软文) and the whole page is not searchable; the same is true with pages 14, 16, and 17. For the same date, on page 2 of the “supplement”, one portion of text is searchable, the other text is not; on page 3 the text for a serial novel is not searchable; on page 5, the cinema column has five articles, but the database classifies it as advertisement and thereby the whole page is not searchable. On March 17, 1936, pages 12, 13, and 14 have several text portions that are not searchable; page 15 is a medical column (an important feature in the newspaper, like the cinema column), but no portion of the page text is searchable.
• The advertisements do not appear to be indexed or searchable in title and/or text (this is very important for economic and cultural studies—commodity, film, entertainment, etc.).