Talk:Suffix tree

Comment in the article history accompanying an edit from [[patricia tree]] to [[radix tree]]

(Terminology: "patricia trie" redirects to "radix tree". Therefore I consider the last term to be the preferred one.)

The edit might be appropriate, but the stated logic is dubious. The most appropriate textual display term should be chosen before interlinking is considered; redirects do not reliably indicate normative usage, rarely distinguish precise usage, and frequently derive from incumbency effects.

This edit should be reviewed for semantic validity. MaxEnt 00:41, 2 June 2006 (UTC)

Important History

This is considered to be an important historical development in the field of algorithms, so I felt it was important to give a fuller account of the historical developments and major contributions. People coming to the article might know of this construction in association with any of the principle innovations, so mentioning each of them helps to contextualize the developments in terms of their prior reference point. MaxEnt 01:20, 2 June 2006 (UTC)

Should be merged with Suffix trie?

Is Suffix trie about the same thing as this? Maybe the title of the other article is misspelt?152.66.129.162 10:48, 16 April 2007 (UTC)

In a suffix tree, all internal nodes are branching. In a suffix trie, they are not, but all edges have length 1. A suffix trie has $O(n^{2})$ nodes, while a suffix tree has $\Theta (n)$. Klem fra Nils Grimsmo 05:52, 19 April 2007 (UTC)

Do you have a reference for using "suffix trie" to mean this separate concept, and for the use of uncompressed tries as a useful data structure? The data structure you describe is useful only pedagogically as a step towards describing suffix trees, as far as I know. I've redirected that article to this one, because I don't see the value of a separate article on a useless data structure. —David Eppstein (talk) 18:12, 3 April 2008 (UTC)

those dashed(fixme) lines are weird a little

at first i though it is not a tree at all :P then i realized they just have to help you. but still i do not see how. those dashes are good. —Preceding unsigned comment added by 84.16.123.194 (talk) 03:21, 24 February 2008 (UTC)

This phrase leaves the reader somewhat hanging:

Suffix trees also provided one of the first linear-time solutions for the longest common substring problem.

This suggests that there are other ways of finding LCS's, maybe discovered even earlier. But the LCS article only lists suffix tries and a matrix oriented solution, which is quadratic. Shinobu (talk) 20:18, 5 October 2008 (UTC)

This is outright incorrect phrase. According to Gusfield, the first linear-time algorithm was proposed by Weiner!! Then this algorithm was improved by MacCreight and Ukkonen.User:itman

Applications

Variants of the LZW compression schemes use it (LZSS).

LZSS is not variant of LZW (LZ78), but LZ77. 83.5.92.106 (talk) 23:53, 8 March 2009 (UTC)