{ 189}

APPENDIX A
GLOSSARY

The following glossary does not cover all jargon terms employed in this book, but only those actually adopted by the book and used repeatedly. Most of the meanings given are based on the work of other authors. The few terms used in an unusual sense in this book are marked with a "+".

access term
in an input string, a term which is the lead term of one or more of the index strings.
ancillary input
in string indexing, any data input to an index string generator other than input strings.
browse
to investigate, without design, the contents of an information store, such as an index or a database.
citation order
the order in which the terms appear within a description such as an index string.
clarity: of an index string, the readiness and likely correctness with which searchers will interpret to what it refers.
code
a meaningful character or sequence of characters which does not contain ordinary human language.
collocation
the extent to which similar elements are close together and dissimilar elements separated (for example, in an index).
connective
a word, phrase, punctuation mark, or other symbol comprehensible to searchers and used to show the relationship between parts of an expression such as an index string.
cross-reference
an element of an index which points to another part of the index.
description
any data, such as a text or other string of characters, which describes or characterizes something else.
effectiveness
of a search, the extent to which it leads to relatively useful as against relatively useless information. { 190}
efficiency
the extent to which less effort is expended to achieve equal effectiveness.
+eliminability: of an index element, the ease with which it can be eliminated by a searcher if it does not contain useful information.
golist
a list of terms which should be access terms.
heading
the first part of an index entry or of any other index element when displayable separately.
index
a kind of retrieval tool, consisting of at least an ordered list of index entries, plus, optionally, other elements such as cross-references.
index display
a display of an index or of part of an index for people to use.
index entry
an element in an index, consisting of a description and a locator.
index string
the description part of an index entry in a string index.
+index string generator
computer software for producing a set of index strings from a single description of an indexed item.
indexed item
a document or other item of information for which a description is produced for purposes of producing an index.
indexer
a person making up descriptions of indexed items for the purpose of producing an index.
input string
in string indexing, a string of terms and other symbols which is a description of an indexed item and which may be used by an index string generator to produce a set of index strings for the indexed item.
keyword
in string indexing, a word in an input string which the index string generator recognizes as an access term.
lead term
the first term in an index string.
level
the length of the sequence of links between a point in a tree and the point designated as the "root".
+link
a direct connection between two terms in a description of an indexed item, corresponding to a relationship between the things that the terms represent.
+linktype
a category to which some links belong.
locator: a string of characters useful in finding the location of information outside an index.
network
of terms and links, a structure in which there is no limitation on which terms may be linked to which.
precision (ratio)
the proportion of the indexed items retrieved during a search on a query, which are judged relevant to that query.
predictability
of an index string, the extent to which a searcher can infer what form it will take and where it will appear in the index if it exists.
preferred term
a term which may be used in descriptions of indexed items
qualification function
the function of leading a searcher from the concept of a more general class to the concept of a subclass without representing the subclass by a new term. { 191}
recall (ratio)
the proportion, of the indexed items judged relevant to a query, which are retrieved during a search on that query.
redundancy
unnecessary repetition; specifically, of information not needed by searchers because they already have it.
relevance
usefulness to an information need; especially, as judged from the expression of the information need in a query.
role code
a code categorizing how a term in a description of an indexed item is connected to other terms in the description or to the indexed item itself.
scan
to examine every element in a set of elements (for example, every subheading under a given heading).
search
in an index, the process of looking for information about things outside the index.
searcher
a person looking for information (for example, from an index or from a database).
stoplist
a list of words, phrases, or other expressions which should not be access terms.
string index
a type of index in which the description part of each index entry is a string of terms and connectives joined together by computer software according to regular and explicit rules and in which each indexed item normally has two or more index entries with overlapping descriptions.
subheading
the part of an index entry or other index element following the heading when displayable separately.
succinctness
the ratio of detail to length of a description such as an index string.
term
a sequence of characters used as the name of a thing or of a class of things; more loosely, any sequence of characters treated in a similar way in indexing.
text
a string of characters constituting a continuous discourse in ordinary language.
thesaurus
a structure indicating terms or other descriptive strings to be used in descriptions of indexed items and the relationships of these to other terms or descriptive strings.
tree
of terms and links, a structure in which there is one and only one sequence of links between any two terms.

<-- Chapter 9: The Future of String Indexing Contents Appendix B: References -->