Abstract
In this work we consider comparison of complex information objects (CIO) as a component of intelligent decision-making. The specifics of proposed approach is that we compare not all theoretically possible CIOs but only their subset that is relevant for cur- rent situation and contains existing and available objects. Thus, we find an acceptable solution from the set of available ones that can be not optimal (according to certain criteria). We propose formal ontology-based model of CIO that considered as an element of intellectual information system. This model (in contrast to the domain ontology) defines unique names for positions of classes and class individuals to indicate the SIO structure. The methods of CIO comparison based on the use of knowledge from the relevant do- main ontology are considered. Various approaches to determining of semantic proximity and semantic similarity are considered as metrics for quantitative evaluation to select parameters of information objects that can be used for calculation of these evaluations. We propose an algorithm for semantic comparison of CIOs which are based on the same ontology and have a similar structure. This algorithm allows generation of comparison criteria and determining hierarchy of this criteria for the current situation. We propose to evaluate the semantic-level similarity of the elements of individual CIOs to certain reference CIO defined by the user (as a description of the optimal solution or generated CIO properties). As a result, a subset of CIOs that satisfy the user requirements is cre- ated, but we have to select only one CIO among them, which will be used in the future to fulfill the user task. Therefore, we need in a set of criteria for CIO comparison and methods to determine the importance of each of these criteria at the current moment in time. For this purpose, we propose to use the method of hierarchical analysis based on a pairwise comparison of the importance of individual criteria.
Publisher
National Academy of Sciences of Ukraine (Co. LTD Ukrinformnauka) (Publications)
Reference32 articles.
1. 1. PRYIMA, S. M.& STROKAN, O. V. & ROGUSHINA, J. V. et al. (2020). Ontological Analysis of Outcomes of Non-formal and Informal Learning for Agro-Advisory System AdvisOnt. International Conference on Technologies and Innovation, Springer, Cham. P. 3-17. Available from: https://link.springer.com/chapter/10.1007/978-3-030-62015-8_1. [Accessed 12/07/2022].
2. 2. KOZYRKOV C. (2019). Introduction to Decision Intelligence. Available from: https://towardsdatascience.com/introduction-to-decision- intelligence-5d147ddab767 [Accessed 12/07/2022].
3. 3. ANISIMOV A.V. & MARCHENKO O.O. & NIKONENKO A.O. (2008). Algorithmic model of associative-semantic contextual analysis of texts in natural language. Problems in Programming. No. 2-3, P. 379-384. Available from: https://core.ac.uk/reader/38468700. [Accessed 12/07/2022].(in Ukrainian).
4. 4. KRYUKOV K.V. & PANKOVA L.A. & PRONINA V.A. and others (2010). Measures of semantic similarity in ontologies. Proc. of Scientific sessions MIFI-2010. Information and telecommunication systems. Problems of information security, vol. 5. S. 75-78. . (in Russian).
5. 5. LASTRA-DÍAZ J. J. & GOIKOETXEA J. & TAIEB M.A.H. et al. (2019). A reproducible survey on word embeddings and ontology-based methods for word similarity: linear combinations outperform the state of the art. Engineering Applications of Artificial Intelligence, 85, P.645- 665. Available from: https://doi.org/10.1016/j.engappai.2019.07.010. [Accessed 11/07/2022].
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献
1. Predicting the Effects of News on the Financial Market Based on Machine Learning Technology;2023 IEEE 5th International Conference on Advanced Information and Communication Technologies (AICT);2023-11-21