@@ -1493,7 +1493,7 @@ are index columns of the multicolumn GiST index.
1493
1493
You can run <acronym>k-NN</acronym> search for all built-in and
1494
1494
custom data types that support the concept of distance. In database
1495
1495
queries, k-NN search can provide performance benefits for
1496
- various use cases: georgraphic information systems (<acronym>GIS</acronym>),
1496
+ various use cases: geographic information systems (<acronym>GIS</acronym>),
1497
1497
classification problems (majority voting), clusterization,
1498
1498
looking for similar objects or nearest events, detecting
1499
1499
duplications and typos, trigram matching, multimedia search, and more.
@@ -1520,9 +1520,9 @@ are index columns of the multicolumn GiST index.
1520
1520
the tuple entry is used. In the tree nodes, leaf nodes are stored before the
1521
1521
inner nodes in the list, so tuple entries are processed first. Each time
1522
1522
a result is returned from the queue, it is guaranteed to be the closest
1523
- one to the target point. Thus, there is no need to scan all nodes.
1524
- Once the required k entries are extracted from the queue,
1525
- search can stop.
1523
+ one to the target point among the remaining entries . Thus, there is no
1524
+ need to scan all nodes. Once the required k entries are extracted from
1525
+ the queue, search can stop.
1526
1526
</para>
1527
1527
</listitem>
1528
1528
0 commit comments