Recall (recall ratio)
Recall is one among other measures of the performance of an information system. It was used for the first time in the so-called "Cranfield II experiments" (Cleverdon et al., 1966). Recall is defined as the proportion between retrieved relevant documents and the total number of relevant documents in the system in question.
| Recall = a : (a + c) X 100%, where a = number of retrieved, relevant documents, c = number of non-retrieved, relevant documents (sometimes termed "silence"). |
Recall is thus an expression of how exhaustive a search for documents is. Recall
constitutes together with precision the most common
measures (and concepts) for the measurements of the performance of retrieval
systems.
In information retrieval exists a long range of strategies for the increase of
recall. Among those are searching in both "natural language fields" as in
controlled vocabulary fields, the inclusion of synonyms (in natural language) in
the query, inclusion of broader terms, restricted use of "and" in Boolean logic
etc. Such strategies are, however seldom without costs but have a tendency to
lower precision.
Recall is difficult to measure on operational systems, why test collections have to be used in research.
Literature:
Cleverdon, C. W.; Mills, J. & Keen, E. M. (1966). Factors determining the
performance of indexing systems. Vol. 1-2. Cranfield, U.K.: College of Aeronautics.
Birger Hjørland
Last edited: 12-02-2007