r10 - 05 Nov 2006 - 12:11:16 GMT - FriederSchmidtYou are here: Bernstein > Main Web >  SystemsIntegration > UseCaseRequirements > UseCaseList > DbSearchAmbiguity

Databases search ambiguity

This is part of the submitted UseCaseList.

Scenario

A description of the scenario that you have in mind.

While searching the user doesn't want to be put before ambiguous choices. Issues - The classification and search systems are problematic for the majority of potential users of the repertories given the number of subjective classification choices which confuse the user. Sources of ambiguity are of different types: search path ambiguity, naming ambiguity...

Importance

How important do you see this use case as?

  • Must Support
  • Ambiguity negatively affects automated searches too.

Dependencies

What other use cases are affected by the implementation of this one?

  • Databases usability / db integration / dating / authentication / cartography / bibliography

Input

Things that the user must/might supply to the system.

  • User looks for a watermark in the databases.

Output

Things that the user will recieve in response to their request.

  • The user found the watermark.

Difficulties

Areas in which you foresee problems/issues arising.

  • The classification and search systems are problematic for the majority of potential users of the repertories given the number of subjective classification choices which confuse the user. Sources of ambiguity are of different types: search path ambiguity, naming ambiguity...

Example

An example supporting this use case.

  • Search path ambiguity:
    1. Level error - Some underclasses are displayed at the same level than the class. For example (POL) "Key" is a top class of the same level as "Tool", yet semantically the key is a tool.
    2. In class "Ochsenkopf" (POL) the user can choose between "Mit Augen", "Mit Augen und Nasenlöchern", "Mit Augen und Maul", "Mit Augen, Nasenform und anderen Gesichtsstrukturen", ... If a user has a wm with an "Ochsenkopf" which has "Augen", "Nasenlöchern" and "Maul" there is nothing the description that would guide him towards one or another of these four classes. Especially confusing is the existence of a path "Ochsenkopf > Mit Augen und Maul > Mit Nasenlöchern" distinct from "Ochsenkopf > Mit Augen und Nasenlöchern".
    3. Heraldic motiv - Rich coat of arms pose a problem since they are classified by the main element inside the shield, while the biggest element by its size - and thus the most likely to the retained for search by users - is outside, encircling the shield. In most cases it's an eagle. The weirdest example is that of a big eagle with two small fishes on his chest classified under "Fish". Other example: the WM "Krone im Adlerherzschild" (POL) is to be found under "Krone" although the WM element with the biggest size is "Adler", and therefore most users would probably search under "Adler". Also the WM are clssified as "Krone im Adlerherzschild", although there is also a "Krone" on top of the "Adler". Same for "Tower > A tower in shield > Escutcheon in fess point of an eagle".
    4. Heraldic class - The classification of heraldic motif seems not to be entierly consistent. For example (POL) "Escutcheon in fess point of an eagle" is found at least under two classes - "Crown" and "Coat of arms > Estucheon per fess > Above florish".
  • Naming ambiguity: POL-124832 is classified as a "unicorn", yet you don't need too much imagination to look for it under "horse". Idem for the class "Flag" which is to be found under "Weapons".

Other Information

Any other information that you think is important to include.

  • ...

Comments

Comment from other partner regarding the use case.

  • Its not always and for for everyone possible to decide this watermark is a lion or a dog. And if the user have only parts of watermarks its becomes still more difficult. Perhaps we need a open search-field such as at google. The user can search in this case for 'animal' + 'dog+lion'. - GeorgDietz

  • These ambiguities exist since the beginnings of scientific filigranology. Cf. Briquet 2179-2192 “Deux Seaux pendus à un écot» = two buckets hanging at a stub; there is no shield in each case, but Briquet presents them among “Armoiries. Seaux”, because of ”Ce filigr. représente les armoires de Galéas II Visconti (1347-1402)” – Briquet knows it is a coat of arms and so you have to look at Armoiries, but no shield gives you a hint that it is a coat of arms. Or look at Briquet 1385: a huge basilisk dominates the appearance of the watermark, but a small shield with a crosse of Basle reigns denotation. – FriederSchmidt

-- VladAtanasiu - 13 Sep 2006

Edit | WYSIWYG | Attach | Printable | Raw View | Backlinks: Web, All Webs | History: r10 < r9 < r8 < r7 < r6 | More topic actions
 
The Bernstein Website
Copyright © by the contributing authors.
Bernstein - The Memory of Paper http://www.bernstein.oeaw.ac.at
Ideas, requests, problems regarding Bernstein? Send feedback
This site is powered by the TWiki collaboration platform