Talk:Word completion

Unusable Material
I removed this material for the reasons stated. meatclerk 05:20, 16 August 2006 (UTC)

Many people believe that word completion is a good idea for predictive text systems. While the idea is appealing in principle, it works quite poorly in practice. Using word completion for general text entry (as opposed to selected from a short list as in the case of selecting urls in browsers) causes confusion and increased mental processing time, which easily outweighs the advantages of reduced keystrokes. Some early work on this subject was done by Dunlop and Crossan, 2000. sometimes word completion does not work properly. Some phones it will not work on are the tracfones because they don't work very well. They don't work very well because they don't work. A citation to the published version of the paper is now included.
 * The artice it references clearly states that it should not be cited.

Word completion is often referred to as tab completion, as in some software such as Bash the completion of a string is invoked by hitting the tab key.
 * Too specific an example to be useful

Examples of word completion in general text editing
I have flagged this section as POV, since it consists almost entirely of unsupported arguments against word completion in general text editing. The Dunlop and Crossan article cited in support of the arguments in fact reports results that are exactly opposite to those claimed in this section: their user tests showed a 10% decrease in the time taken to type words when using a "hybrid word-completion and word-prediction method" on a simulated mobile phone. (I suspect the article can't be accessed without a Springer subscription, so unfortunately this will be difficult for some people to confirm.) Also, that article only reports studies of text entry on mobile phones. It's not clear whether or to what extent their results apply to general text editing. Dunlop and Crossan make no claims in this direction.

The rest of the section presents a one-sided argument against word completion in general text editing, without citing any other references to support the arguments. (The cited Tugboat article discusses average word length, but doesn't make any mention at all of word completion, let alone whether average word length would affect it's usefulness.) The section seems to be a personal point of view on word completion in general text editing. I suggest either undertaking a far wider survey of the relevant literature, or deleting this section entirely.

Finally, apart from the POV problems, the section title would lead one to expect a list of examples of the use of word completion in general text editing, which is not at all what it actually does. (The examples appear in the following similarly-titled section.) 137.222.80.132 (talk) 12:32, 10 December 2008 (UTC)

the example points out that many words begin the same way, and it is not until you've typed 4 letters that you have a reasonable chance of finding the intended, fairly common, word "soccer". Don't see how that is POV. Ipthief (talk) —Preceding undated comment added 00:11, 17 May 2011 (UTC).

Examples of Word Completion- Google
Under Examples of Word Completion, should Google Suggest be added? 69.204.164.119 21:31, 18 March 2007 (UTC)David