| Similarity |
Page |
Snapshot |
| 79 |
level
...Were we to group documents according to the keywords they shared,then for each keyword we would have a group of documents,namely,those which had that keyword in common
...D 1 K 1,K 2 D 2 K 2,K 3 D 3 K 1,K 4 img src Fig
...The usefulness of this kind of structure will become more apparent when we discuss searching of classifications
...Threaded lists In this section an elementary knowledge of list processing will be assumed
...A simple list representation of the classification D 1,D 2,D 3,D 4,D 5,D 6,D 7,D 8 |
| 80 |
is given in Figure 4
...img src Fig
...A modification of the implementation of a list structure like this which makes it resemble a set of ring structures is to make the right hand pointer of the last element of a sublist point back to the head of the sublist
...One disadvantage associated with the use of list and ring structures for representing classifications is that they can only be entered at the top
...Another modification of the simple list representation has been studied extensively by Stanfel [21,22]and Patt A HREF REF
... |
| 100 |
A and B are two clusters
...that their corresponding documents are less dissimilar than some specified level of dissimilarity
...Let us now look at other ways of representing clusters
...where Di is usually the Euclidean norm,i
...More often than not the documents are not represented by numerical vectors but by binary vectors or equivalently,sets of keywords
...remember n is the number of documents in the cluster by the following procedure
... |
| 84 |
eventually terminate at a particular node from which no further branches will emerge
...By now it is perhaps apparent that when we were talking about ring structures and threaded lists in some of our examples we were really demonstrating how to implement a tree structure
...Another example of a tree structure is the directory associated with an index sequential file
...The use of tree structures in computer science dates back to the early 1950 s when it was realised that the so called binary search could readily be represented by a binary tree
... |
| 186 |
behaviour of any one of the components depends in only an aggregate way on the behaviour of the other components
...2
...On the file structure chosen and the way it is used depends the efficiency of an information retrieval system
...Inverted files have been rather popular in IR systems
...There are many more problems in this area which are of interest to IR systems
...3
...So far fairly simple search strategies have been tried
... |
