Where are semantics in semantic web

Title	Where are the Semantics in the Semantic Web?
Journal	AI Magzine
Year	2003
Author	M Uschold
Level	Introductory
Comment	Is an article Semantic's introductory article.

 

 

Summary

 

The most widely accepted Semantic NetworkFeatures

Is machine-usable content. In this sense, semantic networks have matured in terms of shopping agents, because they can automatically obtain and use web information to find the cheapest ticket. But where is the semantics? In fact, most people think that semantic web is still a vision, so such shopping agents should not be counted as semantic web applications. In order to use web content, machines must be 'aware of what they should do when they encounter, which requires machines to be 'aware of their meaning (meaning ).The greatest challenge of semantic Networks

How to make machines possess this knowledge, which is the most confusing place for semantic networks.

 

We start by describing the meaning of "semantics". Different things can be considered to have different types of semantics. We introduced a semantic continuity system, which

The scope is from the implicit semantics (

Implicit
Semantics only exists in the user's brain) to the standard Semantics
(

Formal Semantics, capable
Processed by machines ). We listed some prerequisites (or requirements) for achieving the 'machine available' goal '),

Many problems have been taken into account, such as hard coding, protocols, clarification of semantic conventions, and semantic public declarations. Based on the above requirements and problems, and our semantic continuity, we believe that

Shopping agents is beneficial to regard it as a degraded semantic network, because shopping agents works in a network that lacks semantics. It simply relies on programmers to hardcode some rules into the software to implement its functions.

 

We have noticed that this method has many shortcomings, which provide some ideas for the development of semantic networks.

We believe that this evolution can be performed in the following ways:

 

(1) semantic interval movement: from implicit semantics to standard Semantics

 

(2)
Reduce

Hardwired provides network content Semantics

 

(3) increase the number of protocols and standards

 

(4)
Semantic ing and translation capabilities when development encounters inconsistency

1 Introduction

 

From different aspects, the development of Web includes:

 

Locating resources:

People search for information on the Internet from simple text information and keyword searches to complex semantic searches and browsing.

Users:

Web resources are from the primary consumers to the consumers and machines.

Web tasks and services:

From information retrieval to processing information.

 

The foundation of these capabilities is to make the web have semantic information, which reminds us to look at the development of web from another perspective:

Semantics:

Web information includes a small amount of semantic information, or even semantic information, to a structure rich in semantic information.

 

Although the term "Semantic Web" is widely used, it does not have a widely accepted definition. W3C Semantic Web organization claims that:

 

"The Semantic Web is a vision: the idea of having data on the Web defined and linked in a way that it can be

Used by machines not just for display purposes, but for automation, integration and reuse of data Processing SS

Various Applications. [W3C 2001] "[emphasis mine]

 

Similarly, Tim Berners-Lee expected:

 

"The semantic web is an extension of the current web in which information is given well-defined meaning,

Better enabling computers and people to work in cooperation. "[Berners-Lee et al 2001] [emphasis]

"[S] omething has semantics when it can be 'processed and understood by a computer, 'such as how a bill can

Be processed by a package such as Quicken. "[Trippe 2001]

 

Although "Semantic Web" is not widely accepted, according to the above description, we can find that they mainly emphasize that the Web can:

• Machine-usable, and
• Associated with more meaning

However, we still lack a feature that can clearly distinguish semantic networks from existing networks.

 

Because RDF is adopted by W3C as a semantic Web language, some people think that if an application uses RDF, It is a semantic web application. This assertion reminds us of this sentence: "If a program is written in Lisp or Prolog, It is Ai ".

 

The lack of a clear definition of semantic networks does not prevent us from moving forward. Just like the development history of AI, there is always a period of incubation of new ideas.

 

Research institutes, industrial participants, software developers, and W3C work together to turn the vision of semantic networks into reality. It will be a layered, scalable, and composite architecture. One of the main requirements is to use semantic metadata for expression and reasoning, and provide semantic tagging information for information resources. The foundation of the semantic framework is ontology, knowledge base, and web proxy with the ability of reasoning, proof, and semantic query.

 

The main purpose of semantic web is to enable machines to understand information resources to better serve humans. According to Tim Berners-Lee's vision, web agents will be everywhere in the future, and they will execute a lot of daily work for us. Web agents must communicate with each other and understand each other's meanings to complete their assigned work; they must publish their own functions and identify the functions of other web agents; they also need to locate information resources on the web and combine these resources in a meaningful way to effectively execute tasks. They need to identify, interpret, and respond to the behavior of other web proxies.

 

The two extreme solutions to the above problems are:

1. Ignore this problem. Assume that all proxies use the same vocabulary and have the same meaning.
2. A formal language description ontology that publishes the functions of each agent

 

2. Article Structure

 

1 Introduction

 

2 Semantic: A random-splendored thing

2.1 A semantic continuum

2.1.1 Implicit Semantics

2.1.2 Informal Semantics

2.1.3 Formal Semantics for Human Processing

2.1.4 Formal Semantics for Machine Processing

3 Machine processible
Semantics

3.1 Issues and Assumptions

3.1.1 Language Heterogeneity

3.1.2 Incompatible Conceptualization

3.1.3 Term Heterogeneity and Different Modeling Styles

3.2 An Example

 

4 Why do Web Shopping Agents work?

4.1 Requirements for Machine Usable Content

4.1.1 Hardwiring

4.1.2 Agreements and Semantics Specification

4.1.3 Publicly Declared Concepts

4.2 Web shopping agents work because...

 

5 Conclusion

 

3. Conclusion

 

There are many different ideas about what a semantic network is and how it should be developed. Machine usable Web content is a feature that is often mentioned in semantic Web. It is based on the fact that machines must "know" how to identify the content they are looking for and "know" how to process them, the acquisition of these "Knowledge" requires machines to access the semantic information of web content. But what does this mean? The most confusing way to access the semantic information of web content is in the semantic web. The main goal of this article is to remove this confusion.

 

We started to introduce semantic networks and noticed some major challenges that may be encountered when implementing them. The main problem may be the heterogeneity of web content. We have discussed some meanings of "Semantics", including real-world Semantics, KM semantics and Model Theoretical semantics, and their relationships, we have said that the key to successful semantics is whether the web application can correctly determine the semantic information of the web content. We have noticed many things with semantic information, such as subjective terms and expressions, behavioral languages used for communication between agents, and representation languages (RDF or DAML + oil)

 

How does a machine access the semantic information of web content? To obtain the essence of this problem, we introduce a semantic continuity system. In various cases, the semantics must be hard-coded into web application software. In addition, we need to note that it is in semantic continuity, the last semantics is formal semantics for machine processing. In this case, we can reduce the amount of hard encoding and dynamically infer the semantics of things, however, this requires hard coding to represent the semantic information of a language (which is used for reasoning.

 

We have provided a simple example. This example uses formal semantics. Machines can infer the semantics of the term fuel-pump (not previously seen). This example can be implemented using the current technology, however, we have made some assumptions to eliminate various heterogeneity (language, terminology, modeling style, etc ). An important point in this example is that its implementation relies on a publicly declared term semantic information, which is shared between various software agents and web sites. We also pointed out that such a public statement does not need to be formal; Dublin Core is an example of a non-formal and widely used public statement.