Notifications

376 views

Description

When using text search, a phrase is broken up into it's component words. Matches are then brought back based on the frequency of the words, and whether they were used in sequence. For example, if searching for the phrase "Quick brown fox" your search will be on the words "quick", "brown", and "fox".

In logographic and syllabic writing systems (e.g., Japanese, Chinese), individual characters represent a word/phrase/syllable. Spaces are not used between characters because they aren't necessary for meaning like they are in phonographic writing systems. For example, "quickbrownfox" doesn't make sense because each letter represents a sound rather than a concept, whereas the translation of "Quick brown fox" into Japanese is "速い茶色のキツネ".

If one was searching in English for the text "Crazy quick brown fox" they would expect articles matching "Quick brown fox" to be returned as well (as it contains 3 of the same words). However, the translation of "Crazy quick brown fox" into Japanese is "狂気の速い茶色のキツネ" and because there are no spaces, this "breaking up" of phrases into its components needs to be done on a per-character basis. This is done when searching on $knowledge.do, however it is not done when using contextual search. This means that articles with the text "速い茶色のキツネ" (quick brown fox) are not brought back if the contextual search term is "狂気の速い茶色のキツネ" (crazy quick brown fox).

This difference has an extremely large impact on the usability of contextual search for incident deflection. Unless Japanese users enter an exact phrase into the short description of an incident, relevant matches will not be returned even if they very closely match the phrase.

Steps to Reproduce

Pre-requisites

  • Helsinki Patch 1 instance with demo data
  • I18N: Internationalization plugin installed
  • I18N: Japanese Translations plugin installed

Steps

  1. Navigate to Knowledge > Create New.

     

  2. Enter the following values in the fields and press Submit:
    Knowledge base: IT
    Category: IT > Android
    Language: Japanese
    Short description: 速い茶色のキツネ
    Text: 速い茶色のキツネ
    (NOTE: 速い茶色のキツネ means "Quick brown fox" in Japanese.)

  3. Open the article you just created and press the Publish UI Action.

  4. Navigate to Knowledge > Create New.

  5. Enter the following values in the fields and press Submit:
    Knowledge base: IT
    Category: IT > Android
    Language: English
    Short description: Quick brown fox
    Text: Quick brown fox
  6. Open the article you just created and press the Publish UI Action.

  7. Navigate to sysapproval_approver.list and open the approval records for both articles.

  8. Press the Approve UI Action on each approval record.

  9. Navigate to Incident > Create New.

  10. Type "Quick brown fox" in the Short description field - notice the results that show in the contextual search.

  11. Type "Crazy quick brown fox" in the Short description field - notice the results that show in the contextual search.

  12. Click the gear icon in the upper right corner of the screen and change the systems settings for the session language to Japanese.

  13. Close the System Settings dialog.

    The incident form reloads.

  14. Type 速い茶色のキツネ in the Short description field.

    Note the results that show in the contextual search.

  15. Type 狂気の速い茶色のキツネ in the Short description field.

    Despite adding only the text 狂気の (crazy) before the text 速い茶色のキツネ (quick brown fox), no results are returned although there is a knowledge article with the text 速い茶色のキツネ (quick brown fox) in it.

 

 

Workaround

There is no known workaround.


Related Problem: PRB688183

Seen In

Geneva Patch 3 Hot Fix 2
Geneva Patch 6 Hot Fix 2
Helsinki Patch 1
Helsinki Patch 5

Associated Community Threads

There is no data to report.

Article Information

Last Updated:2017-12-19 17:17:16
Published:2016-07-28