Semantic networks

Starting with the concept molecule we can define a relationship to structure with the relational term consists of. Following the route we can relate structure to substructure, and to functional group. Here we introduce the new relation term defines to describe the relationship between functional group and reactivity. As we follow through all cross-references, we build up a complex picture of the concept of molecule and its relations to other concepts, like biological activity. 

FIGURE 2.2 A semantic network of a molecule describing the relationship between entities with the relational terms consists of or defines and their attributes with the term has. Following the route from the molecule we can relate structure to substructure, functional group, and atom. The atom entity provides attributes, such as partial charge or polarizability, which finally are part of the definition of biological activity. 

Semantic networks have several features that make them particulary useful [3,4]  

Semantic networks represent ontologies, and the inherent knowledge can be retrieved by using automatic reasoning methods [5]. An ontology in computer science is a data model that represents a set of concepts within a domain and the relationships 

Each term requires a unique identifier (id), a name, as well as at least one is a relationship to another term. Additional data are optional in the example, a namespace, its definition, and a synonym. The namespace modifier, for instance, allows the definition of a group in which the term is valid and can also be defined for other modifiers, like the is a relationship. 

---

Traditional symbolic models typically capture the semantic content of symbols through external devices, such as look-up tables and semantic networks. In this approach, symbols serve as pointers to semantic content, rather than capturing that content directly (see Hinton, 1990 Hummel Holy oak, 1997). 

Once acquired, there are many ways of representing the knowledge in the knowledge base, including production rules, frames, semantic networks, decision tables, and trees and objects. Probably the most common methodology is the production rule, which expresses the relationship between several pieces of information by way of conditional statements that specify sections under certain sets of conditions, for example ... 

Process simulation and modelling teehniques are very useful for optimizing design and operation. The outstanding advantage of the knowledge based simulator is its flexibility and semantic network. 

Figure 12.5. Hendler s hybrid model, which joins a neural network and a semantic network (Adapted from Hendler, 1991, with permission from Ablex Publishing Corporation)...

Hendler s (1991) hybrid model combines a semantic network with a neural network, as shown in Figure 12.5. In essence, this model depends upon the neural net to learn the internal representations (i.e., the essential microfeatures) that are associated with a set of input stimuli. Thus, the model develops the hidden unit layer of the network as well as the weights connecting the hidden units to the output units. After the network has settled (i.e., has learned to classify the inputs appropriately), the top two layers of units are accessed by the semantic network model by means of spreading activation. Thus, the nodes of the neural net communicate with the nodes of the semantic net. 

Instead of representing knowledge in a static way, rule-based systems represent knowledge in terms of rules that lead to conclusions. A simple rule-based system consists of a set of if-then rules, a collection of facts, and an interpreter controlling the application of the rules by the given facts. Other important knowledge representation techniques are frames and semantic networks [1],... 

Semantic Networks are a form of knowledge representation in a directed graph, where the nodes represent concepts and the edges, or connectors, describe their relations. 

Shastri, L., Why Semantic Networks in Principles of Semantic Networks Explorations in the Representation of Knowledge, Sowa, J.F., Ed., Morgan Kaufmann Publishers, San Mateo, CA, 1991, 109. 

Lehmann, F., Semantic Networks, Comp. Mathemat. Applicat., 23, 1, 1992. 

Hsing, M. and Cherkasov, A., Integration of Biological Data with Semantic Networks, Current Bioinformatics, 1(3), 1, 2006. 

KL-ONE is a frame-like family of knowledge representation approaches in the tradition of semantic networks [11], It was developed to overcome some of the drawbacks of semantic networks and represents conceptual information in a structured network for inheritance. The frames — in the KL-ONE approach called concepts — are arranged in a class-like hierarchy that includes the relations between frames. Frames are typically inherited from super classes. 

---