IEEE Systems, Man and Cybernetics Magazine - October 2021 - 34

irrelevant, other answers can support with the information
that is also relevant to the query. Figure 2 shows the user
interfaces of IERS with sections to provide input text or
voice to educate IERS.
IERS implementation includes three servers. The user
interface is built with Vue.js framework. A Python server is
used to perform reading and learning from documents.
Another back-end server, built with Node.js, is used to
communicate between the user interface and database,
which is MySQL.
IERS offered the following key functions and features:
◆ reading from documents and capture human experience,
in a structured manner, associated with the
industrial operation related to procedures, tools,
equipment, documents, location, and action mentioned
in the documents
◆ developing a knowledge network, HESN, from the
knowledge it acquires from the document and do reasoning
to produce accurate output relative to the questions
when the IERS is asked
◆ developing a knowledge network, HESN, from the
knowledge it acquires from the document and do reasoning
to produce accurate output relative to the questions
when the IERS is asked
◆ filtering out the search result based on parameters such
as temperature, environment, lab condition, and so on,
which helps to reduce search time and the risk of accident
◆ receiving a query using both text and voice, which is
helpful for operators who require wearing gloves and
at the same time also need to use IERS
◆ training using documents or small texts. It can also be
taught new terms related to the domain knowledge
and it updates HESN and improves its reasoning. The
more it learns, the more it becomes intelligent, and the
more it can serve well.
Conclusion
This article presents a tool, IERS, that learns from documents,
text, or voice and is ready to answer related questions
when asked. It is dependent on the proposed
knowledge structure, called HESN, which captures human
experience from the text or voice input. The system was
tested with 242 textual data; 30 random queries were
asked, out of which 13 were related to specific information
about any instruction or operation, 12 were related to
retrieving list of steps for an operation, and five were related
to possible next or previous step of an instruction. Out
of 30 queries, 27 queries were answered properly, which is
equivalent to 90% accuracy. Table 1 shows where the precision
and recall for nine randomly selected queries are measured.
Although more than one piece of information is
retrieved for each query, the most relevant one is shown to
the user and other ones appear as a list when the user
wants to see more results. Future tasks include improvements
in learning and query processing for better reasoning,
which will result in better accuracy.
34 IEEE SYSTEMS, MAN, & CYBERNETICS MAGAZINE October 2021
Acknowledgments
We would like to thank Ontario Power Generation and the
Natural Sciences and Engineering Research Council for
their support of this research. Thanks also to Matthew
Samson for his support of the development of IERS.
About the Authors
Hossam A. Gabbar (Hossam.Gaber@uoit.ca) is with the
Faculty of Energy Systems and Nuclear Science, Ontario
Tech University, Oshawa, Ontario, Canada, L1G 0C5, USA.
He is the principal investigator of the intelligent experience
retention system.
Sk Sami Al Jabar (sksamial.jabar@ontariotechu.net) is
with the Department of Electrical and Computer Engineering,
Ontario Tech University, Oshawa, Ontario, Canada, L1G 0C5,
USA. He is the research assistant, lead software developer, and
system architect of the intelligent experience retention system.
Hassan A. Hassan (Hassan.Hassan@opg.com) is with
Inspection and Reactor Innovation, Ontario Power Generation,
Whitby, Ontario, Canada, L1N 9E3, USA.
Jing Ren (Jing.Ren@uoit.ca) is with the Department of
Electrical and Computer Engineering, Ontario Tech University,
Oshawa, Ontario, Canada, L1G 0C5, USA.
References
[1] T. Li and Z. Chen, " An ontology-based learning approach for automatically classifying
security requirements, " J. Syst. Softw., vol. 165, p. 110,566, July 2020. doi: 10.1016/j.
jss.2020.110566þ.
[2] C. Wu, P. Wu, J. Wang, R. Jiang, M. Chen, and X. Wang, " Ontological knowledge base
for concrete bridge rehabilitation project management, " Automat. Construction, vol.
121, p. 103,428, Jan. 2021. doi: 10.1016/j.autcon.2020.103428.
[3] E. M. Sanfilippo, F. Belkadi, and A. Bernard, " Ontology-based knowledge representation
for additive manufacturing, " Comput. Ind., vol. 109, pp. 182-194, Aug. 2019. doi:
10.1016/j.compind.2019.03.006.
[4] J. L. Martinez-Rodriguez, I. Lopez-Arevalo, and A. B. Rios-Alvarado, " OpenIE-based
approach for knowledge graph construction from text, " Expert Syst. Appl., vol. 113, pp.
339-355, Dec. 2018. doi: 10.1016/j.eswa.2018.07.017.
[5] B. Xu and H. Zhuge, " The influence of semantic link network on the ability of question-answering
system, " Future Gener. Comput. Syst., vol. 108, pp. 1-14, July 2020. doi:
10.1016/j.future.2020.02.042.
[6] M. Huang, J. Lin, Y. Peng, and X. Xie, " Design a batched information retrieval system
based on a concept-lattice-like structure, " Knowl.-Based Syst., vol. 150, pp. 74-84,
June 2018. doi: 10.1016/j.knosys.2018.02.041.
[7] M. P. Fanti and W. Ukovich, " Decision support systems and integrated platforms:
New approaches for managing systems of systems, " IEEE Syst., Man, Cybern. Mag.,
vol. 3, no. 1, pp. 18-22, Jan. 2017. doi: 10.1109/MSMC.2016.2623868.
[8] A. K. Sharma, S. Chaurasia, and D. K. Srivastava, " Sentimental short sentences
classification by using CNN deep learning model with fine tuned word2vec, " Procedia
Comput. Sci., vol. 167, pp. 1139-1147, Jan. 2020. doi: 10.1016/j.procs.2020.03.416.
[9] M. S. Haleem, L. Han, P. J. Harding, and M. Ellison, " An automated text mining
approach for classifying mental-ill health incidents from police incident logs for datadriven
intelligence, " in Proc. IEEE Int. Conf. Syst., Man Cybern. (SMC), Oct. 2019,
pp. 2279-2284. doi: 10.1109/smc.2019.8914240.
[10] Q. Zhao, " Reasoning with awareness and for awareness: Realizing an aware system
using an expert system, " IEEE Syst., Man, Cybern. Mag., vol. 3, no. 2, pp. 35-38,
Apr. 2017. doi: 10.1109/MSMC.2016.2550013.

IEEE Systems, Man and Cybernetics Magazine - October 2021

Table of Contents for the Digital Edition of IEEE Systems, Man and Cybernetics Magazine - October 2021