Gu H, Perl Y, Geller J, Halper M, and Singh M. A methodology for partitioning a vocabulary hierarchy into trees. Artificial Intelligence in Medicine, 15(1):77--98, January 1999.  @ NUS  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... through descriptions, have focused on specialised applications; those such as Cimino who have allowed uncontrolled multiple hierarchies, have found themselves with hierarchies which are difficult to maintain without transformations analogous to the orthogonal taxonomies advocated here [57]. However, it should be noted that GRAIL is not quite sufficiently powerful to carry through this principle completely in practice. It lacks constructors for the conjunction and disjunction of elementary concepts. As a result high level elementary concepts such as phenomenon must be inserted as ....

H. Gu, Y. Perl, and J. Geller, "A methodology for partitioning a vocabulary hierarchy into trees," Artificial intelligence in Medicine, pp. (in press), 1998.

....the number of levels in the hierarchy is usually limited, since existing codes have a fixed number of digits and each digit indicates a level. Alternatively, some system do not use code to indicate hierarchical location, e.g. Read Version 3, the MED (Medical Entities Dictionary) and SNOMED RT. Gu et al. 1999) proposes a methodology to partition vocabularies into an isA hierarchy. Authors show how to partition an existing MED dictionary, which comprises 48,000 concepts, over 61,000 isA links and over 71,000 additional links (categoryOf, roleOf) Based on the partitioning into sets of concepts with the ....

Gu, H., Perl, Y., Geller, J., Halper, M. & Singh, M. (1999). A methodology for partitioning a vocabulary hierarchy into trees. Artificial Intelligence in Medicine, 15, pp. 77-98.

No context found.

Gu H, Perl Y, Geller J, Halper M, and Singh M. A methodology for partitioning a vocabulary hierarchy into trees. Artificial Intelligence in Medicine, 15(1):77--98, January 1999.

....concepts a network 632 times the size of the excerpt of 68 concepts shown in Figure 1. Each class contains about 477 concepts on average. In Figure 9, we show a large portion of the OOHTR schema s subclass hierarchy, with attributes and relationships omitted by applying information thinning [29, 30]. The figure contains about half of the property introducing classes and the intersection classes. The area classes above the dashed line represent areas rooted at property introducing concepts. Those below the line are intersection classes. Let us point out that intersection concepts may lie at ....

Gu H, Perl Y, Geller J, Halper M, Singh M. A methodology for partitioning a vocabulary hierarchy into trees. Artificial Intelligence In Medicine 1999;15(1):77--98.

....to increase the depth of the Semantic Network. For example (Figure 2 (b) the classes X Y and Z are the only two minimal superclasses of the intersection class X Y Z. Compared with Figure 2 (a) with 5 subclass relationships, Figure 2 (b) contains only 4 subclass relationships. In [14], we defined the complexity of a schema as the ratio between the number of relationships and the number of classes. Thus, when two schemas contain the same classes, the schema with fewer relationships is simpler, and Figure 2 (b) is simpler than Figure 2 (a) Using the above rule to define ....

Gu H, Perl Y, Geller J, Halper M, Singh M. A methodology for partitioning a vocabulary hierarchy into trees. Artificial Intelligence in Medicine, 15(1):77--98, 1999.

....probably not solvable by polynomial algorithms; 2) The requirement of meaningful groups is beyond the scope of graph algorithms. To overcome these problems, we have previously combined human expert judgment with algorithmic tools. We have introduced a technique called disciplined modeling [14, 24] that results in a partitioning of a semantic network into a set of trees. However, in this approach, human expert judgment is an important ingredient. In this paper, we present an approach which completely avoids the involvement of a human expert. Rather, it relies on structural features of the ....

....there are few large trees left. In addition, even the largest of those trees is considerably smaller than the original vocabulary. 3 Structural Partitioning vs. Expert Partitioning In this section, we will compare our partitioning results with the results obtained by semantic partitioning [14]. We apply these two methods to the most complex subnetwork of the MED. The results turn out to be quite similar. In the MED, the concept Cortisporin Opthalmic Ointment has the most ancestors: 39. We will focus on the subnetwork containing this concept and all its ancestors. The subnetwork ....

[Article contains additional citation context not shown here]

Gu, H., Perl, Y., Geller, J., Halper, M., Singh, M.: A methodology for partitioning a vocabulary hierarchy into trees. Artificial Intelligence in Medicine 15 (1999) 77--98

....from Experimental Model of Disease Sign or Symptom Functional Concept to Sign or Symptom and Functional Concept as in Figure 4. Figure 7 shows the alternative modeling. Compared with Figure 4, with 19 subclass relationships, Figure 7 contains only 18 subclass relationships. In [22, 23], we defined the complexity of a schema as the ratio between the number of relationships and the number of classes of the schema. Thus, when two schemas contain the same number of classes, the one with more relationships will be of higher complexity. Hence, the schema in Figure 7 is simpler than ....

Gu H, Perl Y , Geller J, Halper M, Singh M. A methodology for partitioning a vocabulary hierarchy into trees. Artificial Intelligence in Medicine 1999;15(1):77--98.

Online articles have much greater impact   More about CiteSeer.IST at NUS   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST at NUS - Copyright Penn State and NEC. Hosted by the School of Computing, National University of Singapore.