GISdevelopment.net --> Application --> Miscellaneous
Reza Khatami1 (Presenting Person),Ali Asghar Alesheikh2, and Majid Hamrah3
1 Graduate student of GIS, Faculty of Geodesy and Geomatics Engineering,
K. N. Toosi University of Technology,Khatami@sina.kntu.ac.ir
2 Associate professor, Faculty of Geodesy and Geomatics Engineering,
K. N. Toosi University of Technology, Alesheikh@kntu.ac.ir
3 Assistant professor, Faculty of Geodesy and Geomatics Engineering,
K. N. Toosi University of Technology
Abstract
Various geospatial information communities, public authorities and private institutions, recognize the World Wide Web as a medium to distribute their data. Therefore, Web is known as a digital information dissemination media. Such data is required in several stages of decision making process; namely intelligence and design. SDSS can greatly benefit from the information on the web.
With the increasing volume of data on the web, the next challenge that IT professionals are now facing is the use of such data. The use requires mechanisms to understand, search, retrieve, and share data via networks in a machine processable format. Lack of semantics has been identified as the main obstacle for data use.
This article investigates some services that semantic web could offer to SDSS such as improved search (based on semantics of information added to the traditional keyword searching); integrating information from different datasets; and data and message exchanging among systems in understandable manners. Hence, a review of semantic web architecture is presented first. The paper describes the use of ontologies via the Internet in an environmental monitoring system. The main idea is that the members of different information groups access to the meaning of data if they can approach the ontologies that have been developed by those who collected the data. RDF is used to define and structure terms and vocabulary used in an information group. The use of RDF/Schema is described related to the example.
1. INTRODUCTION
Nowadays Web contains a vast amount of information. However, since this information is put into Web through natural languages and mark ups (which are generally used for data representation rather than data retrieval), the semantics of this information are not machine-accessible and one would face lots of problems in information searching, retrieval, combination, and exchanging. Indeed, in virtual environment just the same as our real environment a shared language to transmit concepts and semantics is needed. What Semantic Web tries to do is to make such a shared language that firstly optimizes information access and secondly provides machine processing for information efficiently, by information semantics transmission.
2. CURRENT WEB STATE
The current Web (consisting of HTML language, HTTP protocol, and URI addressing mechanism) causes everyone become able to put his/her information into it. As a result, Web faces an interesting extension. Actually, the aim of Web developers was to make it popular. The current Web has been successful in gathering a vast amount of data, but data search and retrieval are its main deficiencies. Hence, there is no way to search information but using screen-scraping method based on keywords.
3. SEMANTIC WEB
Semantic web was introduced by Burners-Lee on 1998 for the first time, but its formal explanation that has a layered architecture was presented on 2001. Semantic Web is not a separate web from current web; rather it is the future of current web that has lots of enhancements. "The first step is putting data on the web in a form that machines can naturally understand, or converting it to that form. This creates what I call a Semantic Weba web of data that can be processed directly or indirectly by machines."[1] W3C has defined a seven-layered architecture for Semantic Web [http://www.w3.org/2001/sw/]. Figure 1 shows this architecture.
Figure 1- The architecture of semantic web
RDF (Resource Description Framework) is a mechanism that uses directed graphs to represent resources and their attributes. Figure 2 shows a simple RDF graph. Each node of this graph was specified uniquely with an Uniform Resource Identifier (URI). Associating a URI with a resource means that anyone can link to it, refer to it, or retrieve a representation of it [4].
Figure2 – A simple RDF graph (source:http://www.w3.org/TR/2004/REC-rdf-primer- 20040210/)
In the figure, resources are not just web pages and documents, and could be people, concepts, and any thing else.
RDF graphs are composed of triples that have three sections: subject, predicate, and object. For example in "James sells cars" statement James, sell, and car are subject, predicate, and object. RDF Schema (RDFS) is an ontology language that includes concepts like classes, subclass relations, properties, subproperty relations, and restrictions of domain and range. RDF Schema can describe any domain. One of the main usages of RDF Schema is restricting triples that could be expressed in RDF. In such cases, one can define which properties to be used for each class and which values they could get. In RDF Schema hierarchical relationships between classes and inheritance can be defined using rdfs:subclassof in which a class can have multiple superclasses. The same can be done for properties. For example, “is brother of” is a subproperty of “is sibling of”.
3.1. ONTOLOGY
Ontology originates from philosophy and known Aristotle as its source. In philosophy, ontology is a filed of science that studies existences and relations between them. "An ontology is an explicit and formal specification of a conceptualization"[5]. In web contexts, ontologies provide one understanding from a domain. Such an understanding is necessary to solve heterogeneity problem, because, two applications may use two different terms for one concept or use the same term for two different concepts. Actually, ontologies provide semantic interoperability. An ontology defines the common words and concepts (meanings) used to describe and represent an area of knowledge, and so standardizes the meanings. ontologies are used by people, databases, and applications that need to share domain information [2]. Ontology representation languages are called knowledge representation languages. Examples of such languages are RDF/RDFS, DAML+OIL, OWL, Ontolingua, KIF etc.
The expressive power of RDF/RDFS is limited; therefore, W3C has developed OWL (Web Ontology Language) to represent ontologies in web. Ontology language working group established in W3C on 2001 and published the first version of OWL on 2003. OWL is a language to represent ontologies in web that provides their understanding by machine. OWL is mapped from Description Logic that is a subset of Predicate Logic and provides efficient reasoning.
OWL has three levels of language (in the order of expressive power increasing)
In choosing an ontology language, one should consider that there is a trade off between expressive power and efficient reasoning. Then, a compromise between them must be made. Choosing a language as an ontology language depends on expressive power and reasoning power that one needs in his/her domain.
4. SEMANTIC WEB AND SDSS
SDSS can be defined as an interactive, computer-based system designed to support a user or group of users in achieving a higher effectiveness of decision making while solving a semi-structured spatial decision problem [6]. The structured part of the problem may be amenable to automated solution by the use of a computer, while the unstructured aspects are tackled by decision makers [7].
Each SDSS has three basic parts: a GIS database, a MCDM (Multi Criteria Decision Making) system, and a user interface. The problems typically require the integration of very large volumes of data and information from various sources and the coupling of this information with efficient tools for assessment and multi-criteria evaluation that allow broad, interactive participation in the decision–making process [8]. There are two main approaches to combine these parts: "Loose coupling" and "Tight coupling". Apart from the approach used, a common sense between these parts is needed to exchange data that sometimes requires export / import tools. Semantic Web can create such a common sense to make relationships between these parts more efficiently. One of the main problems that each SDSS faces with is the combining various datasets from various resources. However, this problem can be solved without semantic web, but it is a time and energy consuming task. Since documents on Semantic Web are machine processable, then a main part of this task can be done by machine. Semantic web brings this ability to SDSSs to combine knowledge and data from various datasets. The meaning of combined data is accessible with not only human but also machines and agents. Therefore, SDSSs could access to a more extensive knowledge and data on the web and could make decisions that are more reliable.
As we can see from the figure 3, we have extended an architecture for a SDSS web service. In this architecture, the user firstly introduces the ontologies to the system and then the system transforms the different ontologies into our desired architectures by an ontology parser like "Hewlett Packard's Jena." The next step is the combination of these ontologies by ontology alignment engine through the following methods:
Finally, the SDSS engine receives the combined ontologies. SDSS is made up of a GIS Toolbox and a MCDM, which do the spatial processes and decision-making analysis respectively. The interrelation between these two parts is based on the combined ontology. SDSS can interact with the users through a user-interface, get the required parameters, and finally represent the results.
Figure 3 – The Architecture of a SDSS Web Service
5. DECISION MAKING AGENTS
SDSS agents can survive on web, but it needs to combine information from various resources on web instantaneously and use different web services that provide decisionmaking tools. These tasks are almost impossible without semantic web, because many problems like data searching, data retrieval and combination, searching web services and many more arise.
Regarding data searching, semantic web improves search (based on semantics of information added to the traditional keyword searching) and about combining datasets, it facilitates this issue by using ontologies and ontology alignment techniques. Another obstacle to reach to the decision-making agents is searching web services. However, web services are described by standards like WSDL (Web Services Description Language) and UDDI (Universal Description, Discovery, and Integration) and are searchable, but in near future, when they increase, it will be difficult to find them. This problem is the result of the fact that these standards are not able to transmit semantics and when an agent faces a web service just syntax can be shared. This serious problem is more paralyzing when the agent needs to combine and orchestrate some web services. Semantic web can provide web services and agents interaction in an understandable manner.
6. GSDSS
GSDSS (Group SDSS) is a special kind of SDSS in which decision makers are distributed in space and time. It is an interactive, computer-based system that facilitates the solution of unstructured problems by a group of decision makers [9]. It aims to improve the process of group decision making by removing common communication barriers, providing techniques for structuring decision analysis and systematically directing the pattern, timing, and content of the discussion [10].
In addition to the principal components of conventional MC-SDSS (data and analysis module, MCDA module, and interface), a MC-GSDSS should contain communication capabilities and enhances the model base to provide voting, ranking, and rating for developing a consensus [6].
It is observed that the prevalent problems are of the same kind similar to the enterprise SDSS ones with a difference regarding their solvings, which is more vital and complicated in their cases. Therefore, there is a greater demand toward Semantic Web. Added to what mentioned, GSDSSs could use semantic Web technologies to receive and process different inputs from different professionals with distinct ontologies and to provide understandable outputs and maps for each group depending on their ontologies and vocabularies, which makes GSDSSs more flexible.
7. Conclusion and Recommendations
Today SDSS, because of its needs, face lots of difficulties naming data searching, data combination, knowledge management, and etc. These difficulties can get more complex when data and decision-makers are distributed. In a situation that SDSS is an agent, resolving these obstacles is believed to be almost impossible. Throughout this paper, first Semantic Web is introduced and then presenting the current SDSSs' obstacles, we tried to investigate the role of Semantic Web to overcome these barriers.
Interested readers are recommended to work on ontology alignment techniques to solve heterogeneity problems in spatial domain and also querying from aligned ontologies. Utilizing GIS context knowledge can enhance the results of these alignments and queries.
REFERENCES:
How semantic web can influence SDSS
Reza Khatami1 (Presenting Person),Ali Asghar Alesheikh2, and Majid Hamrah3
1 Graduate student of GIS, Faculty of Geodesy and Geomatics Engineering,
K. N. Toosi University of Technology,Khatami@sina.kntu.ac.ir
2 Associate professor, Faculty of Geodesy and Geomatics Engineering,
K. N. Toosi University of Technology, Alesheikh@kntu.ac.ir
3 Assistant professor, Faculty of Geodesy and Geomatics Engineering,
K. N. Toosi University of Technology
Abstract
Various geospatial information communities, public authorities and private institutions, recognize the World Wide Web as a medium to distribute their data. Therefore, Web is known as a digital information dissemination media. Such data is required in several stages of decision making process; namely intelligence and design. SDSS can greatly benefit from the information on the web.
With the increasing volume of data on the web, the next challenge that IT professionals are now facing is the use of such data. The use requires mechanisms to understand, search, retrieve, and share data via networks in a machine processable format. Lack of semantics has been identified as the main obstacle for data use.
This article investigates some services that semantic web could offer to SDSS such as improved search (based on semantics of information added to the traditional keyword searching); integrating information from different datasets; and data and message exchanging among systems in understandable manners. Hence, a review of semantic web architecture is presented first. The paper describes the use of ontologies via the Internet in an environmental monitoring system. The main idea is that the members of different information groups access to the meaning of data if they can approach the ontologies that have been developed by those who collected the data. RDF is used to define and structure terms and vocabulary used in an information group. The use of RDF/Schema is described related to the example.
1. INTRODUCTION
Nowadays Web contains a vast amount of information. However, since this information is put into Web through natural languages and mark ups (which are generally used for data representation rather than data retrieval), the semantics of this information are not machine-accessible and one would face lots of problems in information searching, retrieval, combination, and exchanging. Indeed, in virtual environment just the same as our real environment a shared language to transmit concepts and semantics is needed. What Semantic Web tries to do is to make such a shared language that firstly optimizes information access and secondly provides machine processing for information efficiently, by information semantics transmission.
2. CURRENT WEB STATE
The current Web (consisting of HTML language, HTTP protocol, and URI addressing mechanism) causes everyone become able to put his/her information into it. As a result, Web faces an interesting extension. Actually, the aim of Web developers was to make it popular. The current Web has been successful in gathering a vast amount of data, but data search and retrieval are its main deficiencies. Hence, there is no way to search information but using screen-scraping method based on keywords.
3. SEMANTIC WEB
Semantic web was introduced by Burners-Lee on 1998 for the first time, but its formal explanation that has a layered architecture was presented on 2001. Semantic Web is not a separate web from current web; rather it is the future of current web that has lots of enhancements. "The first step is putting data on the web in a form that machines can naturally understand, or converting it to that form. This creates what I call a Semantic Weba web of data that can be processed directly or indirectly by machines."[1] W3C has defined a seven-layered architecture for Semantic Web [http://www.w3.org/2001/sw/]. Figure 1 shows this architecture.
Figure 1- The architecture of semantic web
RDF (Resource Description Framework) is a mechanism that uses directed graphs to represent resources and their attributes. Figure 2 shows a simple RDF graph. Each node of this graph was specified uniquely with an Uniform Resource Identifier (URI). Associating a URI with a resource means that anyone can link to it, refer to it, or retrieve a representation of it [4].
Figure2 – A simple RDF graph (source:http://www.w3.org/TR/2004/REC-rdf-primer- 20040210/)
In the figure, resources are not just web pages and documents, and could be people, concepts, and any thing else.
RDF graphs are composed of triples that have three sections: subject, predicate, and object. For example in "James sells cars" statement James, sell, and car are subject, predicate, and object. RDF Schema (RDFS) is an ontology language that includes concepts like classes, subclass relations, properties, subproperty relations, and restrictions of domain and range. RDF Schema can describe any domain. One of the main usages of RDF Schema is restricting triples that could be expressed in RDF. In such cases, one can define which properties to be used for each class and which values they could get. In RDF Schema hierarchical relationships between classes and inheritance can be defined using rdfs:subclassof in which a class can have multiple superclasses. The same can be done for properties. For example, “is brother of” is a subproperty of “is sibling of”.
3.1. ONTOLOGY
Ontology originates from philosophy and known Aristotle as its source. In philosophy, ontology is a filed of science that studies existences and relations between them. "An ontology is an explicit and formal specification of a conceptualization"[5]. In web contexts, ontologies provide one understanding from a domain. Such an understanding is necessary to solve heterogeneity problem, because, two applications may use two different terms for one concept or use the same term for two different concepts. Actually, ontologies provide semantic interoperability. An ontology defines the common words and concepts (meanings) used to describe and represent an area of knowledge, and so standardizes the meanings. ontologies are used by people, databases, and applications that need to share domain information [2]. Ontology representation languages are called knowledge representation languages. Examples of such languages are RDF/RDFS, DAML+OIL, OWL, Ontolingua, KIF etc.
The expressive power of RDF/RDFS is limited; therefore, W3C has developed OWL (Web Ontology Language) to represent ontologies in web. Ontology language working group established in W3C on 2001 and published the first version of OWL on 2003. OWL is a language to represent ontologies in web that provides their understanding by machine. OWL is mapped from Description Logic that is a subset of Predicate Logic and provides efficient reasoning.
OWL has three levels of language (in the order of expressive power increasing)
- OWL Lite
- OWL DL (Description Logic)
- OWL Full
In choosing an ontology language, one should consider that there is a trade off between expressive power and efficient reasoning. Then, a compromise between them must be made. Choosing a language as an ontology language depends on expressive power and reasoning power that one needs in his/her domain.
4. SEMANTIC WEB AND SDSS
SDSS can be defined as an interactive, computer-based system designed to support a user or group of users in achieving a higher effectiveness of decision making while solving a semi-structured spatial decision problem [6]. The structured part of the problem may be amenable to automated solution by the use of a computer, while the unstructured aspects are tackled by decision makers [7].
Each SDSS has three basic parts: a GIS database, a MCDM (Multi Criteria Decision Making) system, and a user interface. The problems typically require the integration of very large volumes of data and information from various sources and the coupling of this information with efficient tools for assessment and multi-criteria evaluation that allow broad, interactive participation in the decision–making process [8]. There are two main approaches to combine these parts: "Loose coupling" and "Tight coupling". Apart from the approach used, a common sense between these parts is needed to exchange data that sometimes requires export / import tools. Semantic Web can create such a common sense to make relationships between these parts more efficiently. One of the main problems that each SDSS faces with is the combining various datasets from various resources. However, this problem can be solved without semantic web, but it is a time and energy consuming task. Since documents on Semantic Web are machine processable, then a main part of this task can be done by machine. Semantic web brings this ability to SDSSs to combine knowledge and data from various datasets. The meaning of combined data is accessible with not only human but also machines and agents. Therefore, SDSSs could access to a more extensive knowledge and data on the web and could make decisions that are more reliable.
As we can see from the figure 3, we have extended an architecture for a SDSS web service. In this architecture, the user firstly introduces the ontologies to the system and then the system transforms the different ontologies into our desired architectures by an ontology parser like "Hewlett Packard's Jena." The next step is the combination of these ontologies by ontology alignment engine through the following methods:
- Terminological methods
- Structural (internal and external) methods
- Extensional (based on instances) methods
- Semantic methods (based on models)
Finally, the SDSS engine receives the combined ontologies. SDSS is made up of a GIS Toolbox and a MCDM, which do the spatial processes and decision-making analysis respectively. The interrelation between these two parts is based on the combined ontology. SDSS can interact with the users through a user-interface, get the required parameters, and finally represent the results.
Figure 3 – The Architecture of a SDSS Web Service
5. DECISION MAKING AGENTS
SDSS agents can survive on web, but it needs to combine information from various resources on web instantaneously and use different web services that provide decisionmaking tools. These tasks are almost impossible without semantic web, because many problems like data searching, data retrieval and combination, searching web services and many more arise.
Regarding data searching, semantic web improves search (based on semantics of information added to the traditional keyword searching) and about combining datasets, it facilitates this issue by using ontologies and ontology alignment techniques. Another obstacle to reach to the decision-making agents is searching web services. However, web services are described by standards like WSDL (Web Services Description Language) and UDDI (Universal Description, Discovery, and Integration) and are searchable, but in near future, when they increase, it will be difficult to find them. This problem is the result of the fact that these standards are not able to transmit semantics and when an agent faces a web service just syntax can be shared. This serious problem is more paralyzing when the agent needs to combine and orchestrate some web services. Semantic web can provide web services and agents interaction in an understandable manner.
6. GSDSS
GSDSS (Group SDSS) is a special kind of SDSS in which decision makers are distributed in space and time. It is an interactive, computer-based system that facilitates the solution of unstructured problems by a group of decision makers [9]. It aims to improve the process of group decision making by removing common communication barriers, providing techniques for structuring decision analysis and systematically directing the pattern, timing, and content of the discussion [10].
In addition to the principal components of conventional MC-SDSS (data and analysis module, MCDA module, and interface), a MC-GSDSS should contain communication capabilities and enhances the model base to provide voting, ranking, and rating for developing a consensus [6].
It is observed that the prevalent problems are of the same kind similar to the enterprise SDSS ones with a difference regarding their solvings, which is more vital and complicated in their cases. Therefore, there is a greater demand toward Semantic Web. Added to what mentioned, GSDSSs could use semantic Web technologies to receive and process different inputs from different professionals with distinct ontologies and to provide understandable outputs and maps for each group depending on their ontologies and vocabularies, which makes GSDSSs more flexible.
7. Conclusion and Recommendations
Today SDSS, because of its needs, face lots of difficulties naming data searching, data combination, knowledge management, and etc. These difficulties can get more complex when data and decision-makers are distributed. In a situation that SDSS is an agent, resolving these obstacles is believed to be almost impossible. Throughout this paper, first Semantic Web is introduced and then presenting the current SDSSs' obstacles, we tried to investigate the role of Semantic Web to overcome these barriers.
Interested readers are recommended to work on ontology alignment techniques to solve heterogeneity problems in spatial domain and also querying from aligned ontologies. Utilizing GIS context knowledge can enhance the results of these alignments and queries.
REFERENCES:
- Berners-Lee, T. 2002. Weaving the Web. San Francisco: Harper San Francisco.
- Michael C. Daconta,Leo J. Obrst,Kevin T. Smith. "The Semantic Web: A Guide to the Future of XML, Web Services, and Knowledge Management". Wiley publishing.
- Berners-Lee, T., J. Hendler, and O. Lassila. 2001. “The Semantic Web.” The Scientific American. May 2001
- Shadbolt N. , Hall W. ,Berners-Lee T." The Semantic Web Revisited"
- Gruber, T. 1993. “A Translation Approach to Portable Ontology Specifications.” Knowledge Acquisition 5: 199–220.
- Malczewski, J.1999. "GIS and Multicriteria Decision Analysis." John Wiley and Sons
- Hopkins,L.(1984).Evaluation of methods for exploring ill-defined problems. Environment and planning B 11 (3);339-348
- Armstrong, M.P. , P.J. Densham, and K. Kemp (1996). Report From the specialist Meeting on Collaborative Spatial Decision Making, Initiative 17, National Center for Geographic Information Analysis, University of California-Santa Barbara, September 17- 21, 1995.
- DeSanctis,G., and R.B.Gallupe (1987). A foundation for the study of group decision support systems. Management science 33 (5); 589-609
- Sprague,R.H.,Jr., and H.J.Watson (1996). Decision support for management. Upper Saddle River, NJ: Prentice Hall.