Group:
naming
Topic:
manual indexing
Topic:
non-hierarchical classification and multiple classification
Topic:
problem of classifying information
Topic:
problems with information retrieval
Topic:
selecting command names for a user interface
| |
Summary
Language is dependent on context. Users can not predict descriptions of events and entities. There is much overlap, but little agreement about how people assign names. We "stretch and modify meaning" (cbb 11/07)
Subtopic: stretch and modify meaning 
| Quote: in ordinary communication, the ability to stretch and modify meanings is essential [»kentW_1978]
| | Quote: most specifications require verbal reinforcement to resolve ambiguities and questions [»joneTC4_1979]
| | Quote: people distinguish classes via short descriptions; using language specific to those classes; brevity is achieved without lose of information [»bongM_1967]
|
Subtopic: words and phrases used to discuss a subject
| Quote: users can not predict the words and phrases used to discuss a subject, but they think they can [»blaiDC3_1985]
| | Quote: an accident called an event, difficulty, subject of your last letter etc. [»blaiDC3_1985]
| | Quote: three most popular verbs for each operation totaled a third of all responses [»landTK7_1983]
| | Quote: the chance of people using the same content word for the same object ranged for 7% to 18% over a variety of stimulus domains [»furnGW3_1982]
| | Quote: choosing the most common object for a word matches about half of the time [»furnGW3_1982]
|
Subtopic: problem of agreement
| Quote: common concepts may be treated in different ways by different domains; e.g., different currencies [»ahmeR12_1991]
| | Quote: if two scientists judge relevance of a document, only 60% agreement [»clevC4_1984]
| | Quote: when two people search a database, only 40% of output is common
| | Quote: schema conflicts occur when the same information uses different structures, names, data types, or constraints [»kimW12_1991]
| | Quote: the subject similarity among pairs of cited and citing documents is frequently small; they share only one or two topics of the many topics covered by either document [»hartSP10_1993]
|
Subtopic: ambiguous words
| Quote: words are ambiguous and a document can be relevant without using words from the query; concepts more important [»krovR4_1992]
| | Quote: people can often disambiguate words with only a few words of context, and frequently, only one word is needed [»krovR4_1992]
| | Quote: developing a lexicon for disambiguation is expensive; e.g., 8 person-years for 5,000 words [»krovR4_1992]
| | Quote: query words are more ambiguous than words in a document [»krovR4_1992]
|
Subtopic: indexing and thesaurus
| Quote: only 30% agreement when indexing a document
| | Quote: when two people construct a thesaurus only 60% agreement
|
Subtopic: communication between indexer and readers
| Quote: problems aggravated in computerized data bases since fewer conversations about the data [»kentW_1978]
| | Quote: in a graphical browser for Hypertext, the author's interpretation of link icons should match the users' interpretations [»frisME7_1988]
| | Quote: indexers should get feedback from inquirers but usually get feedback from other indexers [»blaiDC_1990]
| | Quote: information retrieval is a process of communication between inquirers and indexers; i.e., a problem of language and meaning, or use [»blaiDC_1990]
| | Quote: want inquirers and indexers to be using search terms in substantially the same way
|
Subtopic: data model
Quote: the data model of a library may use separate objects for two different books, or two copies of the same book, or a single copy at different times
[»holtAW_1979]
|
Related Topics
Group: naming (32 topics, 789 quotes)
Topic: manual indexing (19 items)
Topic: non-hierarchical classification and multiple classification (16 items)
Topic: problem of classifying information (42 items)
Topic: problems with information retrieval (51 items)
Topic: selecting command names for a user interface
(15 items)
|
