Impact of Covid-19 Outbreak on Performance of Indian Banking Sector Ambrish Kumar Mishraa, Archana Patelb and Sarika Jainc a Office of DGA (Energy) Comptroller and Auditor General (CAG), Delhi, India b Institute of Computer Science, Freie University Berlin, Berlin, Germany c Department of Computer Applications, National Institute of Technology Kurukshetra, Haryana, India Abstract The COVID-19 pandemic adversely impacted various industrial sectors of India as well as other countries across globe. In India, impact is resulting to a negative growth rate in economy. Many sectors were performing good before the pandemic but now they have been pulled down by this pandemic. So, it is very much required to analyze and cater the data about those sectors which are badly impacted by pandemic, these sectors play vital role in Indian economy. One of the most important sector of Indian economy is banking sector which is responsible for all the financial activities going on in the country and working as a supporting hand to all of the industries in term of financing, credit, transactions, collection and payment and so on. There are so many reports containing numerous data are in public domain stating the effects of this virus pandemic. The data is not only in physical form but also it is scattered in various format over the internet. Though the data amount is enormous, the major problem is to get the appropriate data according to the user needs. The databases available online are being regularly updated but these databases are not able to provide inference over the knowledge already stored. By using inference capability, we can fetch latent and indirect information out of the knowledge base. Various ontologies for Covid-19 are available online but they do not focus on the performance of banking sector of India during Covid- 19. So, many times users do not get appropriate information according to the imposed query. This article attempts to highlight the repercussions of the Covid-19 in the performance of the Indian banking sector by creating and evaluating the largest comprehensive knowledge base called ontology (Covid19-IBO) in order to get semantic information, in continuation of the same we address few important research questions with respect to Indian economy. Keywords Large data, Ontology, Indian Banking, Covid-19, Sectors, Evaluation 1. Introduction work. Apart from it the new phase of Indian Banking resembles in work like providing forex support, digital banking, e-commerce, telebanking, Indian economy basically depends on the three e-kiosk and many more. You cannot imagine rapid sectors namely primary sector, secondary sector growing economy without banking support. If and tertiary sector and all the three sectors are banking sector get impacted by any obstacle its being majorly supported by banking sector. consequences will definetly be borne by all these Banking sector is providing the financial three sectors which are pillar of the Indian economy. support to all these sectors by disbursing loans, This pandemic appeared as “black swan event” advances, short term credits, issuing letter of that needs immediate action from government to credit, bank guarantees etc as its traditional help resume economic stability through banking ISIC’21: International Semantic Intelligence Conference, February channel [1]. Based on approximation about 25-27, 2021, New Delhi, India recovery time from this global pandemic various ambrishkm.comm@cag.gov.in (A. K. Mishra); arch- anamca92@gmail.com (A. Patel); jasarika@nitkkr.ac.in (S.Jain) economic tools are pointing out towards global ©️ 2021 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). economic depression of different dimensions. CEUR Workshop Proceedings (CEUR-WS.org) Covid-19 has affected the economy of India at that time when the growth rate of the country was at lowest in last 10 year. In the recent past, Indian economy was trying to get on the track by Now, Banks have to minimize the risk and use the recovering with a slow rate. However, due to high risk-averse strategy to restructure loans, this pandemic the recovery process is severely provisioning bad debts due to less risk appetite, impacted. As in last two quarters India has Indian banks have already suffered severe losses in facing negative growth in GDP. The Indian past restructuring attempts. The same report economy was already suffering even before the indicates that 19 sectors are been adversely Covid-19 outbreak, but Covid-19 outbreak impacted by this pandemic resulting the stress of resulting it worsen more. In a recent report dept having value Rs 15.5 lakh crore which were not published by the RBI (India’s central bank) under the stress before this virus outbreak [2]. Fig 1 states that this virus has impacted better (a) and (b) shows the adverse impact on % share in companies, organizations and businesses that banking sector debt and debt in Rs lakh crore were performing well before this pandemic. respectively. Therefore, investigation of the impact of Covid- 19 from the large amount of distributed data is very vital to prevent the downfall of the economy and the minimize the pandemic effect. It is also essential because this study will be used as a touch bearer in future if any of the pandemic impacts like Covid-19. This paper offers the Covid19 impact on Banking ontology (Covid19-IBO) that provides semantic information about the impact of the Covid-19 on the banking sector of India. The major contributions of the paper are listed below: • Development of Covid19 Impact on Banking ontology (Covid19-IBO) • Evaluation of the Covid19-IBO by different evaluation approaches The rest of the paper is divided into six sections. Section 2 describes existing work. Section 3 discusses some research questions that is handled by (a) developed ontology. Section 4 shows the development and evaluation of the Covid19-IBO. Section 5 emphases the result and discussion of the proposed work. Section 6 shows the results of subjective testing and last section concludes the paper. 2. Literature Covid-19 pandemic adverse impact the Indian economy. To control the flow of the virus, GoI announced a nationwide lock down and various policies to help the people. Dev and Sengupta [3] have analyzed the economic condition of the India before the Covid-19 along with policies that has been declared so far and potential effect of the shock on several part of the Indian economy. Rakshit and Basistha [4] have wrote an article about economic effect of the outbreak in India by considering (b) outbreak as a man-made disaster i.e. human tragedy. Figure 1: (a) % share in banking sector debt They addressed three important research questions: (b) debt in Rs lakh crore (Source: data taken the effect of Covid-19 on the Indian economy along from [2]) with the detailed analysis of the different sectors that suffered from Covid-19, the effect of Covid-19 on the bilateral trade relationship between China literature, we claim that the available databases and and India, the performance of health system ontologies that provide information according to the during this pandemic. Kanitkar [5] user queries do not have the complete information demonstrated the economic loss of India during about the impact of Covid-19 on Indian banking Covid-19 by using a linear I/O model and results sector that play vigorous role in the growth of Indian shows that the loss is about 10-30% of its GDP. economy. The author has also focused on the emission of CO2 from the power sector and electricity 3. Research Questions supply, demand. Demirguc-Kunt et al. [6] have analyzed the effect of the Covid-19 outbreak on the banking sector by discussing the bank stock By the current article, three important research prices all over the world along with examine the question are addressed that are listed below: role of financial policy by using global databases for the performance of bank stocks. RQ1. What are the necessary steps to minimize the The Covid-19 data is available on the loss to banking sector by Covid-19 pandemic? internet in various format. WHO provides multilingual Covid-19 database that updates In current situation by the cause of outbreak of this regularly and contains all the information about virus pandemic, Indian banks need to review the Covid-19 [7-8]. Kousha and Thelwall [9] portfolio in asset and liability side, for all the provided the access of the coverage of scholarly discussed cases to easily grasping the negative databases and impact indicators from the period effect. This present economic situation warns more of 21.03.2020 to 18.04.2020 so that people can stress evaluation that might show straight identify the important new studies quickly from implications for settlement that are make by Indian Covid-19 publications like news, tweets, Banks for current time. Finding the high-risk citations, facebook, databases and many more sectors/areas/corporates/individuals and re- places. To respond effectively to emergencies evaluating the credit risk provisions related with like public health, we need to share the loan for various economical cases is inevitable. information across various disciplines and IT systems [10]. This is the place where ontologies RQ2. What are the major challenges for Indian offer excellent services and overcome the banking sector during Covid-19? problem of interoperability. Along with the databases, various ontologies also have been During the virus pandemic, major production units developed in order to exact the hidden and were closed or partially working. Entertainment, semantic information. Dutta and DeBellis [11] Aviation, Tourism industries are badly impacted. have published the ontology as a data model Due to the same liquidity in market needs to namely COviD-19 ontology for case and patient increase with keeping an eye to lowering down the information (called CODO) on the web as a NPA. This is major challenge for banking industry. knowledge graph that provides the information For the same, RBI has infused liquidity of about 3.2 about the Covid-19 pandemic. The primary % of GDP in the system. Now banks can lend to focus of the CODO ontology is to describe the reconstruct / support badly impacted industries but Covid-19 cases and Covid-19 patient data. taking a lighter risk. Right now, many SMEs and Infectious Disease Ontology (called IDO) is an MSMEs are bound to shut down their operations, interoperable ontology that contains the domain surely it is indicating towards increasing of loan information about infectious disease where default cases. Though as a cushion RBI has allowed entities are related to the clinical and biomedical moratorium period, it is not enough to meet the aspects of the disease [12]. The extension of the requirement of industries. Hence RBI has to make IDO and Virus Infectious Disease ontology all efforts to meet the challenges and take the (VIDO) is called COVID-19 Infectious Disease banking industry in right direction. Ontology (known as IDO-COVID-19) and contains the information about the Covid19 RQ3. How much Indian banking sector prepared disease and SARS-CoV-2 virus [13]. for effects of Covid-19 on economy? The available different format of data (text documents, video, audio, databases and Preparation of Indian banking sector lies on ontologies) contains the detailed information sustainability of this virus in a long-time span, it about the Covid-19 disease. After studying the also depends on nature as well as intensity of the shocks given to economy. In current pandemic (Covid19-IBO) represents the accurate knowledge time, future of banking sector depends mainly in the domain of banking sector of India and the on designing the policy and their knowledge storage is complete, continuous, easily implementation now. RBI's proactive approach accessible and readable. Covid19-IBO contains a and stabilizing role is need of hour. However, structured knowledge of Indian banking sector. This RBI mainly tried to reduce the repo rates and knowledge is reusable according the need of user. increase the liquidity in economy. By pumping Collection of data in a structured manner is an fund in the banking system without proactive integral phase of the ontology development process. assessment fiscal measures for fulfilling the For the development of the Covid19-IBO ontology, demands will surely contribute in increasing the we collect the data from different sources (as NPAs. The latest report of CRISIL indicates that mentioned below) according to the need of the banking industry will suffer with an increase of domain. 11.5% in bad loans by March next year. This • Research articles of conferences, journals and may cause uncertainty in banking industry book chapters resulting “discouragement in consumption as • Existing ontology repositories/portals like Bio well as the invested corpus required for pushing portal, EMBL-EBI, Agro portal quick recovery in economy.” We may have a • Articles on websites (As Wikipedia, blogs, dif- huge surplus of money due to supply of funds to ferent sites and so on.) maintain the liquidity and this large corpus may • Covid-19 Databases and ontologies remain underutilized by the households • Conducting interviews with expert person like organizations and industries due the pandemic doctor and general public. effect. How long we face the same situation in • Covid-19 Reports, GoI future? Only time knows, but it is astonishing whether, without demand being assessed, these An ontology is organized in hierarchal manner policies will ever be enough, though they have where every entity is attached with other entities in some relaxing effects in short run. a parent-child relationship. Data property and object property are the properties contained in ontology to 4. Development of Ontology define the relationship between the entities. Use of data property is to relate any individual to a user- This section focuses on the ontology defined value while use of object property is to re- development and its evaluation followed by key late the individuals belonging to any class [16]. information about data collection. Covid19-IBO provides light-weight scope of repre- sentations that reflects various important and re- quired concepts about the impact of Covid-19 on In- 4.1. Creation of ontology dian Banking sector and relations among them. It has knowledge such as schedule and non-schedule Ontology is a knowledge representation scheme bank that can further classified into Indian banks that encode knowledge in the form of classes, and Foreign banks in the form of concepts with re- relationships, properties or features or attributes, lationships like disjoint, operating etc. By using this instances or individual and axioms [14]. information, machines become more capable in in- Ontology helps to infer the implicit domain ferring much practical and relevant knowledge of knowledge as like human does, which cannot be past, present and future situation, that will help peo- achieved using conventional Database, using ple by strengthening their decision-making process domain specified rules. In case of database, we in a strategic and tremendous manner. will define all the relations manually for every Covid19-IBO stores real-world information instance. Therefore, Database does not provide into RDF format (Subject-Object-Predicate) that re- hidden information automatically until it is not duce the storage space as compared to other repre- defined. Although rules are also available in the sentation scheme. Figure 2 (a), (b) and (c) shows the database but it is limited whereas redundancy in classification of concepts, object and data properties RDBS can be reduced using normalization respectively. All object and data properties have do- process but cannot be removed and ontology main and range for example: Axis Bank has_open provides 0% redundancy due to its hierarchical Saving Account (Object Property), PNB Bank nature. Therefore, ontology is used in various has_status NPA(Data Property). applications to structured real-life data [15]. Covid19 Impact on Banking Ontology 4.2. Evaluation Evaluation of the Covid19-IBO is very important in order to know the quality and content of the ontology. The evaluation of Covid19-IBO ontology determines that ontology is well built and comprises all essential concepts and relationships which are required for the reasoning. Evaluation approaches are categories into two group namely Verification that means Building an ontology correctly and Validation that means building the correct ontology. For the verification, we use Quantitatively (Metric based Approach) and Qualitatively (Criteria based Approach) approaches. To validate the ontology, we use competency questions. We have utilized the Pellet and Hermit reasoners (available in the protégé tool) to check the consistency of the developed Covid19-IBO. • Quantitatively: We evaluate the ontology quantitatively by using OntoMetric tool [17] which is a metric-based approach. It divides the (a) (b) features into five metrics namely Schema, Instance, Base, Graphs, and Individual axioms. Table 1 indications the important magnitude of some metrics of Covid19-IBO that shows the richness of the ontology. These magnitudes have been calculated with the help of OntoMetric tool. Table 1 Value of Metrics of the Covid19-IBO Metrics Value Axioms 313 Logical axioms count 137 Class count 105 Object property count 28 Data property count 43 Properties count 71 SubClassOf axioms count 94 SubObjectPropertyOf axioms count 18 SubDataPropertyOf axioms count: 19 Attribute richness 0.409524 Inheritance richness 0.895238 Relationship richness 0.229508 Axiom/class ratio 2.980952 Class/relation ratio 0.860656 (c) • Qualitatively: We evaluate the ontology Figure 2: Covid19-IBO ontology (a) Classes qualitatively by using Ontology Pitfalls Scanner (b) Object Properties (c) Data Properties (OOPs) tool [18] which is a criteria-based approach. OOPs shows the pitfalls or error by determining Consistency, Conciseness, Completeness, Correctness and Clarity of Disease Ontology (IDO-COVID-19) do not contain the Covid19-IBO under the three categories the complete information about the impact of namely minor (not very serious pitfall), Covid-19 on Indian banking sector. important (not critical pitfall but it is important to correct it) and critical (need to remove this pitfall) pitfalls. Figure 3: Pitfalls that has been removed Figure 3 shows all the pitfalls that are Figure 4: Matric value of ontologies namely calculated by OOPS! tool [18] and these Covid19-IBO, CODO, CIDO and IDO-COVID-19 pitfalls need to be removed from the ontology. We have removed all the minor Figure 4 shows the value of different metrics of and major pitfalls of the Covid19-IBO available ontologies as compared to Covid19-IBO. ontology by enhancing it. Covid19-IBO has Although in some cases available ontologies have no critical pitfalls as shown in figure 3. more matric values as compared to Covid19-IBO yet our developed ontologies are sufficient in order • Competency Questions: We validate the to provide the complete and precise information ontology by writing competency questions about each question which are related to the impact that provides better understanding of the of Covid19 on Indian banking sector. Figure 5 scope and objectives of the Covid19-IBO depicts the number of common entities of available and helps to build the correct ontology ontologies as compared to Covid19-IBO. according to the user’s need. Some selected competency questions are mentioned below: a) What is state wise rise in NPA during Covid-19 pandemic? b) How many corporate accounts has been closed during Covid-19 pandemic? c) What is sector wise depreciation rate in credit repayment during Covid-19 pan- demic? d) Which bank and sector suffered most by Figure 5: Number of common entities of Covid19- this pandemic? IBO of available ontologies e) How many bank employee’s loss their job during pandemic? f) How many new accounts, loans, credit 6. Subjective Testing cards are being issued from starting of the pandemic? To test our approach, we have adopted subjective testing suggested by industry experts. Several 5. Result and Discussion important parameters namely Adoption of properties (as compared to existing ontologies), The available ontologies namely an ontology for User friendly (How easily user can adopt), Future collection and analysis of Covid-19 data use (How ontology can be used in future viz (CODO), Coronavirus Infectious Disease vaccination process), Relevance of current time Ontology (CIDO), and COVID-19 Infectious (Referring to the current situation of Covid-19 pandemic), Benefit of ontology, Impact on The developed Covid19-IBO ontology has been economy and Impact on society (How this evaluated by three approaches namely competency ontology can be used by students, research questions, OntoMetric tool and OOPs pitfall scholars, professionals, Industry personnel etc) scanner that shows completeness of the Covid19- has been considered for the rating of this IBO. We have also compared the Covid19-IBO with ontology by the users on the scale 1 to 10 where other accessible and available ontologies of 1 stands for not at all satisfying these parameters Covid19 and infer that the existing ontologies are and 10 for best possible satisfaction. not able to provide complete and concrete answer of Thirty-nine users participated in the the question about the Indian banking sector. subjective testing by filling the google form. The future work of this paper is focus on the Among the thirty-nine users, the eighteen are development of the widget that offer semantic female and twenty-one are male. The average services like visualization, annotation, mapping etc. age of the participant is 28.05 whereas We will also publish the Covid19-IBO on linked minimum and maximum age of participants are open data (LOD) for better accessibility and 21 and 52 respectively. The participated users reusability. belong to different occupation like faculty, engineers, students and scholars etc. Users have also provided additional comments over the References developed ontology like very useful, incorporate all the facts of the Indian economy, ontology is very much required in order to [1] The Economic Times, 2020. URL: analyze the impact of Covid-19 on Indian https://bfsi.economictimes.indiatimes.com/blo banking sector and so on that shows accuracy g/potential-implications-of-covid-19-on-the- and importance of Covid19-IBO ontology. The banking-sector/4227. user’s results are shown by the figure 6 on the [2] Hindustan Times, 2020. URL: basis of the different parameters that suggested https://www.hindustantimes.com/india- by experts. news/70-of-banking-sector-debt-affected-by- covid-19-s-impact/story- MAYiYZWz5NE6Pijm7XQNSJ.html. [3] S. M. Dev, R. Sengupta, Covid-19: Impact on the Indian economy, Indira Gandhi Institute of Development Research(2020), Mumbai April. [4] B. Rakshit, D. Basistha, Can India stay immune enough to combat COVID‐19 pandemic? An economic query. Journal of Public Affairs (2020), 20(4), p.e2157. [5] T. Kanitkar, The COVID-19 lockdown in India: Impacts on the economy and the power sector. Global Transitions (2020), 2, 150-156. [6] A. Demirguc-Kunt, A. Pedraza, C. Ruiz- Figure 6: Ranking of Covid19-IBO Ortega, Banking sector performance during the covid-19 crisis (2020). 7. Conclusion and Future Work [7] WHO Coronavirus disease (COVID-19) situation reports, 2020. URL: The positive economic growth of any country https://datasetsearch.research.google.com/sear reflects the financial soundness and increased ch?query=covid19&docid=g3oiDuHtkLygNk purchasing power of that country. The Covid-19 WHAAAAAA%3D%3D pandemic destroyed the growth of the various [8] Eular Covid-19 database, 2020. URL: economic activities in countries so India as well. https://www.eular.org/eular_covid19_databas Banking sector plays vital role in supporting the e.cfm economy of the country by maintaining [9] K. Kousha, M. Thelwall, COVID-19 liquidity. In order to know the impact of Covid- publications: Database coverage, citations, 19 on the Indian banking sector, we have readers, tweets, news, Facebook walls, Reddit presented Covid19-IBO ontology by analyzing posts. Quantitative Science Studies, (2020), 1- existing data that available in different format. 28. Google Form Link: http://bit.ly/3pP2h1j [10] S. Babcock, L.G. Cowell, J. Beverley, B. [15] S. Jain, A. Patel, Smart Ontology-Based Event Smith, The Infectious Disease Ontology in Identification. In 2019 IEEE 13th International the Age of COVID-19 (2020) Symposium on Embedded Multicore/Many- [11] B. Dutta, M. DeBellis, CODO: An core Systems-on-Chip (MCSoC) (2019), pp. Ontology for Collection and Analysis of 135-142. IEEE. Covid-19 Data, (2020) arXiv preprint [16] A. Patel, U. K. Yadav, S. Jain, Non-monotonic arXiv:2009.01210. Reasoning for Scenario Awareness over [12] BioPortal, 2020. URL: Emergency Knowledge Base. In Proceedings https://bioportal.bioontology.org/ontologie of ICETIT 2019 (2020), pp. 482-489, Springer, s/IDO Cham. [13] OLS Ontology Search, URL: [17] OntoMetrics, URL: https://www.ebi.ac.uk/ols/ontologies/idoc https://ontometrics.informatik.uni- ovid19 rostock.de/ontologymetrics/ [14] D. Kumar, A. Kumar, M. Singh, A. Patel, [18] OOPS!, URL: http://oops.linkeddata.es/ S. Jain, An online dictionary and thesaurus. Journal of Artificial Intelligence Research & Advances, (2019), 6(1), 32-38.