Knowledge technology through functional layered intelligence
Introduction
We are facing intensifying problems with complexity, value determination, and accessibility of information. The problems span multiple aspects of existence and use of information, such as: the interconnectedness of information system, the flow of information across those systems, and the storage or representation of information within a system:
- •
The complexity and levels of interconnectedness of information systems is increasing exponentially.
- •
The amount of information flowing across the global networks is increasing exponentially.
- •
Information is being stored in disparate forms across disparate systems.
This wide spectrum of the problem domain compounds the effects we feel. Companies are struggling with enterprise applications and systems integration, and sharing information within their company as well as with partners. End users are struggling with imprecise web searches and navigation through piles of information. The worldwide collection of information is becoming a network of rapidly growing islands of information, which do not fall under any standard governance or universally effective translation systems. The power to extract useful knowledge from this network is lagging behind its growth significantly.
For decades now, there have been many ongoing efforts to combat these issues through data, information, and knowledge technology but these attempts have yielded very limited success [19]. We have solid data technology, decent information technology, but knowledge technology has not achieved significant applied results. “We are drowning in information and starving for knowledge” [16]. This paper identifies three core limitations to improving knowledge technology:
- 1.
Amorphous and ambiguous elementary concepts.
- 2.
Limitations of the formalist–reductionist approach and the need for organic attributes.
- 3.
Lack of interoperability between existing efforts.
The problem with amorphous and ambiguous elementary concepts relates to problems in defining concepts of human cognition, such as data, information, and knowledge. There is an existing hierarchy in organizational sciences that describes such concepts, referred to as DIKW (Data–Information–Knowledge–Wisdom). DIKW is also known as knowledge hierarchy, information hierarchy, knowledge pyramid, and other synonyms [1]. However, the hierarchy remains nearly as amorphous and ambiguous as its first published reference in T.S. Eliot’s poetry [8] in 1934. Today, the same hierarchy is used in cognitive psychology, knowledge management, organizational theory, and other sciences that work with human cognition. The hierarchy is sometimes altered, such as by Ackoff who adds Understanding between Knowledge and Wisdom [1], or Zeleny, who adds Enlightenment after Wisdom [18]. In all the DIKW formulations, the concepts and the transitions between its elements remain intangible. Knowledge is often de-contextualized, reified, or approached as an attribute to information. As Ikujiro Nonaka [13] points out, Western theories on knowledge “lack the view on the fundamentals of epistemology: what is knowledge, the nature of knowledge, and what constitutes learning. They are not clear about how the knowledge is captured, created, leveraged, and disseminated” [14].
Formalism and reductionism have proven very useful for elementary and practical computation, but they impose rigid requirements of absolute predictability, exactness, and liner execution (even with recursion, loops, jumps, and other flow control mechanisms). With very complex systems, and particularly ones that need to operate in environments that are not precisely defined or predictable, the formalist–reductionist methods begin to struggle. We should also recall that Gödel’s theorems put an end to the search of the universal formal language of nature.
Biological systems on the other side seem to thrive in complexity. Even though slower in their elementary units of execution, they achieve much greater parallelism and interconnection densities that any technological systems. They are empowered with organic attributes such as fault tolerance, self-adaptation, and self-motivation. These organic attributes are sought today for the technological systems. However they are not native to the formalist–reductionist methods.
Lack of interoperability, the third limitation of knowledge technology today, refers to disparity of definitions and implementations of the structural components of knowledge technology, such as data types, schemas, and ontologies. It reflects a lack of a common architectural model in knowledge technology. It can be compared to the inability of the Internet to develop into an effective solution until it adopted TCP/IP as the common architectural model.
The Internet and the WWW are major enablers and frequent components of knowledge technology. The issues they face are a subset of the issues identified above. Hai Zhuge, the chief scientist of China’s Knowledge Grid projects, identifies four limiting characteristics of the Internet and the WWW: rapid expansion of resources and users, micro-structured and macro-less organization of resources, inequality of information, distribution, and machine-unreadable semantics [21]. These issues can be mapped to the issues and limitations of knowledge technology. The rapid expansion of resources and users is the driving factor of the discussed intensifying problem with the complexity, value, and accessibility of information. The micro-structured and macro-less organization and the inequality of information and distribution, relate to the lack of interoperability between knowledge technology efforts. Machine-unreadable semantics also relates both to the lack of interoperability and to the amorphous and ambiguous elementary concepts. Zhuge also points out the necessity of organic characteristics [22].
Section snippets
Knowledge technology today
The interconnectivity of global networks and the Internet has presented unprecedented opportunities as well as challenges as their complexity increases. These challenges have stirred up considerable research projects in knowledge technology. Nonetheless, as Gräther notes, knowledge technology is still in its infancy, as “Current approaches to knowledge sharing communities like recommender systems or shared ontologies often suffer from an imbalance of effort versus benefit from the individual
Functional layered intelligence defined
FLI is an architecture reference model, not a new system or a new set of protocols. FLI is not addressing a lack of protocols or designs, but the unstructured and uncoordinated development of knowledge technology without an overarching functional intelligence paradigm.
The goal of FLI is not to model the human-being in all its intelligence and other mental functions. Instead, FLI is designed to enable integration of human-like intelligence functions into technological systems for specific
A FLI example
The following is a sample FLI layer stack with functions defined in the layers.
Knowledge technology through FLI: Ontologies and knowledge
Understanding the difference between knowledge and information, and providing the proper transformation functions between them is essential to knowledge technology, and probably one of the main holdbacks in its progress.
Knowledge is more than passive sets of information that share a vocabulary. It is the product of their co-relationships and the network they form. Knowledge becomes active when that network is put in use. FLI clearly differentiates between information and knowledge. Ontology, on
Conclusion
The current implementations of the Knowledge layer functions fail to create effective “interconnection environment” needed to generate on-demand ontologies and manage their dynamics and transformations. Many prerequisites for the interconnection environment are not clear themselves. For example, it is still not clear how the semantics of the complex ontology inter-relations (inheritances and combinations) in the Semantic Web will look like [30]. As Zhuge points out, the interconnection
Martin Dimkovski is a Network Manager and an Instructor at Barry University. He leads the Network Operations Center for Barry University. Dimkovski has full time experience in the IT industry since 1998 and holds leading IT certifications. As an Instructor he teaches and designs courses in internetworking. Dimkovski has extensive background in algorithm design and problem solving techniques since 1992, including international mathematical Olympiads. Dimkovski has an MS in IT with a focus in AI,
References (22)
Resource space model, Its design method and applications
Journal of Systems and Software
(2004)From data to wisdom
Journal of Applies Systems Analysis
(1898)- N. Boer, P.J. Van Baalen, K. Kumar, An activity theory approach for studying, in: Proceedings of the 35th Hawaii...
- D. Beckett, RDF/XML Syntax Specification (Revised), Retrieved June 2005, from http://www.w3.org/TR/rdf-syntax-grammar,...
- et al.
The semantic web
Scientific American
(2001) - CKG — China Knowledge Grid Research Group, Retrieved July 2005, from...
- Cycorp Inc. (CYC), Retrieved 14 June 2005, from...
- DARPA’s Real-World Reasoning (REAL) Project Home Page, Retrieved 20 June 2005, from...
The Rock
(1934)- et al.
Creating, Maintaining, and Integrating Understandable Knowledge Bases
(2001)
Community support and awareness enhancements for cooperative knowledge generation
Euromicro
Cited by (5)
OMAS - A flexible multi-agent environment for CSCWD
2011, Future Generation Computer SystemsCitation Excerpt :However, although AADE provides additional functionalities, the basic approach remains the same, the overall system becoming more complex. Dimkovski and Deeb [8] qualify such approaches as formal-reductionist, missing the capacity of implementing real human-like intelligent functions based on interpretation of ontologies. We completely agree with their position and present a platform facilitating the development of systems positioned at the knowledge level as defined in their Functional Layered Intelligent model, but at the same time having the capability of including humans seamlessly.
Current state of knowledge technology in construction
2019, Advances and Trends in Engineering Sciences and Technologies III- Proceedings of the 3rd International Conference on Engineering Sciences and Technologies, ESaT 2018Applications of Knowledge Technology in Construction Industry
2018, ICETA 2018 - 16th IEEE International Conference on Emerging eLearning Technologies and Applications, ProceedingsHow rumors spread and stop over social media: A multi-layered communication model and empirical analysis
2015, Communications of the Association for Information SystemsAgent-based user personalization using context-aware semantic reasoning
2008, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Martin Dimkovski is a Network Manager and an Instructor at Barry University. He leads the Network Operations Center for Barry University. Dimkovski has full time experience in the IT industry since 1998 and holds leading IT certifications. As an Instructor he teaches and designs courses in internetworking. Dimkovski has extensive background in algorithm design and problem solving techniques since 1992, including international mathematical Olympiads. Dimkovski has an MS in IT with a focus in AI, a BS in Computer Science, and going into his Ph.D. He also has extracurricular education and work done in sociology and mathematics. His areas of interest are: AI, nature inspired computing, consciousness, dynamic and network systems.
Khaled Kevin Deeb is an Associate Professor and the Academic Coordinator of Information Technology in the School of Adult and Continuing Education at Barry University. Dr. Deeb joined Barry University on September 1, 1999. Prior to that, he has accumulated a substantial amount of practical experience in the Computer Science field due to his research commitment with NASA during his pursuit of the Doctor of Philosophy in Computer Science and engagement in consulting services in the areas of networking and databases. He has published a number of papers in different tracks of cooperation technologies, databases, software architecture, and artificial intelligence. He was given the best paper award in the area of Artificial Intelligence in SCI 2003.