Knowledge Organization and Representation under the AI Lens

doi:10.2478/jdis-2020-0002

摘要/Abstract

Abstract:

Purpose: This paper compares the paradigmatic differences between knowledge organization (KO) in library and information science and knowledge representation (KR) in AI to show the convergence in KO and KR methods and applications.

Methodology: The literature review and comparative analysis of KO and KR paradigms is the primary method used in this paper.

Findings: A key difference between KO and KR lays in the purpose of KO is to organize knowledge into certain structure for standardizing and/or normalizing the vocabulary of concepts and relations, while KR is problem-solving oriented. Differences between KO and KR are discussed based on the goal, methods, and functions.

Research limitations: This is only a preliminary research with a case study as proof of concept.

Practical implications: The paper articulates on the opportunities in applying KR and other AI methods and techniques to enhance the functions of KO.

Originality/value: Ontologies and linked data as the evidence of the convergence of KO and KR paradigms provide theoretical and methodological support to innovate KO in the AI era.

Key words: Knowledge representation, Knowledge organization, Artificial Intelligence, Paradigms

. [J]. Journal of Data and Information Science, 2020, 5(1): 3-17.

Jian Qin. Knowledge Organization and Representation under the AI Lens[J]. Journal of Data and Information Science, 2020, 5(1): 3-17.

图/表 5

参考文献 29

1	Bench-Capon T.J.M.(1990). Knowledge representation: An approach to Artificial Intelligence London: Academic Press An approach to Artificial Intelligence. London: Academic Press.
2	Berners-Lee T., Handler J., & Lassila O. (2001). The Semantic Web. Scientific American,Featured article. DOI: 10.1038/scientificamerican0501-34
3	Brachman R.J., &Levesque H.J., (Ed.). (1985). Readings in knowledge representation. Los Altos, CA: Morgan Kauffman Publishers.
4	Castano S., &Varese G., (2011). Trust-based techniques for collective intelligence through folksonomy coordination. In Next Generation Data Technologies for Collective Computational Intelligence, ed. Nik Bessis and Fatos Xhafa, Chapter 4, 87-112. Berlin: Springer-Verlag.
5	Chen M., Liu X.Z., & Qin J. (2008). Semantic relation extraction from socially-generated tag: A methodology for metadata extraction. In Proceedings of the Dublin Core International Conference, Berlin, Germany.
6	Davis R., Shrobe H., & Szolovits P. (1993). What is a knowledge representation? AI Magazine, 14, 17-33. Retrieved from
7	Doerr M., Stefan G., Steffen H., Antoine I., Carlo M., &van de Sompel , H. (2010). The Europeana Data Model (EDM). In World Library and Information Congress: 76th IFLA General Conference and Assembly 10-15 August 2010, Gothenburg, Sweden.
8	Feibleman J.K. (1954). Theory of intergrative levels, The British Journal for the Philosophy of Science., 5(17), 59-66
9	Fisher D.H. (1987). Knowledge acquisition via incremental conceptual clustering. Mach Learn 2, 39-172.
10	Gene Ontology Consortium. (2019). The Gene Ontology and the scientific literature.
11	Gruber T.R.(1993). A translation approach to portable ontologies. Knowledge Acquisition, 5(2), 199-220https://doi.org/10.1006/knac.1993.1008
12	Haug P.J. (1993). Uses of diagnostic expert systems in clinical care. Proceedings. Symposium on Computer Applications in Medical Care, 379-383.
13	Hayes P.J. (1974). Some problems and non-problem in representational theory. In Readings in Knowledge Representation, ed. Ronald J. Brachman and Hector J. Levesque. Los Altos, CA: Morgan Kaufmann.
14	Hjørland B. (2008). What is knowledge organization (KO)? Knowledge Organization, 35(2), 86-101. DOI: 10.5771/0943-7444-2008-2-3-86
15	Hoffmann A. (1998). Paradigms of Artificial Intelligence: A methodological & computational analysis. New York: Springer.
16	Humphrey S. (1989). MedIndEx system: medical indexing expert system. Information Processing & Management, 25(1), 73-88.
17	Ingwersen P. (1994)Polyrepresentation of information needs and semantic entities elements of a cognitive theory for information retrieval interaction. In: Croft B.W., van Rijsbergen C.J. (eds) SIGIR ’94. Springer, London.
18	ISO. (2013). Information and documentation - Thesauri and interoperability with other vocabularies. Part 2: Interoperability with other vocabularies. Geneva, Switzerland:ISO.
19	Kubat M., Pfurtscheller G., & Flotzinger D. (1994). AI-based approach to automatic sleep classification. Biol. Cybern. 70, 443-448.
20	Lagoze C., &Hunter J., (2001). The ABC ontology and model. International Conference on Dublin Core and Metadata Applications, pp. 160-176. Retrieved from
21	NCBI. (2018). Taxonomy.
22	Qin J., Yu B., &Wang L.Y. (2018). Knowledge node and relation detection. In: Networked Knowledge Organization Systems (NKOS) Workshop at the Dublin Core International Conference DC-2018, Porto, Portugal, September 13, 2018.
23	Qin J., &Zou N., (2017). Structures and relations of knowledge nodes: Exploring a knowledge network of disease from precision medicine research publications. Proceedings of iConference 2017 (pp. 56-65).
24	Qin J., &Paling S., (2001). Converting a controlled vocabulary into an ontology: The case of GEM. Information Research, 6(2). .
25	Qin J. (2002. Evolving paradigms of knowledge representation and organization: A comparative study of classification, XML/DTD, and Ontology. Proceedings of the Seventh International Society for Knowledge Organization Conference, July 10-12, 2002, Granada, Spain, 465-471. Würzburg, Germany: Ergon.
26	United States. President's Science Advisory Committee. (1963). The responsibilities of the technical community and the government in the transfer of information: A report of the President’s Science Advisory Committee.
27	Toms E. (2019). Artificial intelligence and information science? Discussion thread from ASIST Open Forum.
28	Turing A.M. (1952). The chemical basis of morphogenesis. Philosophical Transactions of the Royal Society of London B, 237(641), 37-72.
29	Yang Y ., &Chute C.G., (1994). An application of Expert Network to clinical classification and MEDLINE indexing. Proceedings. Symposium on Computer Applications in Medical Care, 157-161.

Paradigm		Goals	Methods	Functions
KO	Integration	Organize the knowledge universe	Categorize, classify generalize, synthesize	Represent knowledge in publications and organize knowledge about nature and/or society
KO	Disintegration	Organize the knowledge in a domain	Categorize, classify, generalize, synthesize, model	Represent knowledge in data and publications in a domain
KR	Production rules	Represent knowledge in condition-action pairs to solve problems	Use forward chaining algorithms to execute condition-action pairs	Represent fragmentary knowledge in entity-attribute-value (triple) format
	Semantic networks and frames	Represent semantic relations between concepts	Express semantic relations in triples	Connect knowledge nodes through attributes or slots to form a knowledge graph
	First-order logic	Formalize qualifier construction in natural language	Express declarative propositions using the first-order logic syntax and semantics	Produce a set of axioms for reasoning

Category	Atomic level (name of things)	Concept level	Cluster level
Gene	Her2, BRCA1, BRCA2, EGFR	Oncogenes	EGFR mutations in lung cancer
Disease	Non-squamous carcinoma, squamous cell carcinoma	Non-small cell lung cancer	Lung cancer
Drug	Pertumzumab, Lmatinib, Crizotinib	Tyrosine kinase inhibitor	Oncogene de-addiction

Relationship	Pattern	Example
has-biomarker	Disease has-biomarker Gene	chronic myeloid leukemia has-biomarker BCR-ABL non-small cell lung cancer has-biomarker EGFR
is-driver-of	Gene is-driver-of Disease	Her2 is-driver-of breast cancer c-Kit is-driver-of chronic granulocytic leukemia
targets	Drug targets Gene	Crizotinib targets ALK Olaparib targets BRCA1/2
has-role-of	Drug has-role-of Treatment	Crizotinib has-role-of oncogene de-addiction Olaparib has-role-of DNA repair