In this section, we review related researches in two main categories: 1) methods expressing the application of ontology in financial markets and 2) existing knowledge sources related to financial markets.
2.1 Application of ontology in financial markets
In this field, according to current knowledge, relatively limited researches have been published, which are introduced below.
In the research [3] in order to measure the impact of various types of news on business activities in financial markets, such as trading price fluctuations, trading volume, frequency of trading, an ontology is presented for news related to financial instruments, which consists of two parts: The first part presents a hierarchical framework for domain knowledge, which primarily includes news classes, market participants classes, financial market classes and basic relations between these classes. In the second part, a causal map is used to show how classes of news are related to areas of these markets. In this research, the process of making an ontology (to prove the concept) was done in a completely manual and limited way using the news related to terrorist attacks and its impact on the capital markets. Finally, it has been concluded that first, understanding the knowledge of news in financial fields is a very complex and comprehensive process. Second, building this knowledge base helps build trading models based on news in financial instrument markets. Third, developed systems such as news based stock price prediction systems, support systems for financial market participants for searching relevant news are facilitated and supported by ontology.
In a research [4] has been done in order to gain explicit knowledge from the news. Based on the general mechanism of domain knowledge, knowledge reasoning and knowledge discovery, this paper develops a framework for discovering implicit knowledge from news and applying knowledge in stock forecasting. At first, according to the characteristics of the financial field and the conceptual cube, the conceptual structure of the industry-company-product is constructed and the ontological framework is proposed in a limited field (polyester industry). Second, by constructing the ontology of the financial field, the financial news knowledge management framework is proposed. In addition, using the ontology features and domain rules extracted from the news text, a financial news knowledge reasoning mechanism is built to achieve financial news knowledge discovery, and a rule-based model for stock price prediction is presented, which is claimed to have a proper performance in empirical analysis of the polyester industry.
In the research [5], it is stated that in the field of financial securities and transactions, there is a lack of work to develop ontologies, especially for stock markets and fraud detection. This paper attempts to address this gap by providing a systematic framework for developing and managing an ontology for stock markets. To achieve this goal, researchers use techniques to gather and integrate domain knowledge from existing sources alongside new knowledge obtained from unstructured source analysis based on stock market fraud cases reported by the Securities and Exchange Commission. The proposed framework and knowledge are built through six case studies that evaluate work in different types of stock market maintenance and monitoring programs.
In the research [Zdraveski et al., 2017] a data model has been developed for a stock market. This model is based on a built ontology for the stock market that covers the entities and relationships in the stock market and its stakeholders. The article, which is written very briefly, states that the main purpose of using this ontology is to connect specific stock market and stock transaction data with general company information from DBPedia, and it is expected that merging these two datasets will lead to more accurate statistics or new indicators that help traders see a better big picture of companies and stocks being traded.
In the research [7] it is stated that the news related to the stock market in the last decade has a vital role for brokers or users and therefore the focus of this article is on predicting the sentiments of the stock market news based on their polarity and textual information using a Convolutional Neural Network (CNN), which is based on ontology knowledge. The details of this ontology are not mentioned in the article, but apparently the ontology is built based on users' information, which affects the polarity of opinions.
The paper [8] describes the design procedure of an ontology-based marketplace for trading web services. Based on the analysis of the requirements to create a web service market, a market is designed that has these advantages: The market uses a communication language based on ontology, which is able to represent semantics of described requests, offers, and agreements. In addition, semantic information has been used to divide the entire market into several independent submarkets. This concept is shown to be more efficient than the existing mechanism, it can be said that he uses of background knowledge has reduced the overall complexity.
In the research [9] it is stated that the stock market depends on various factors such as price, volume, etc. that these factors can be easily identified by indicators and also in the literature, there are various indicators available that can be used to reduce risk in the stock market within investing or trading. However, none of the indicators alone can provide complete protection in capital risk. Therefore, it is necessary to either create a new index or propose a framework that consists of a set of several indices that can reduce the investor's risk compared to any of the existing indices. The aim of this paper is to propose an ontology-based framework that simultaneously presents the characteristics of three indicators, i.e., Bollinger Bands, Fibonacci Bands, and Heikin-Ashi. In this paper, these indicators are mapped with an ontology in the form of classes, features, and examples.
According to the reviewed researches and the confirmation of the importance of knowledge in identifying factors and characteristics influencing the price prediction process or the direction of stock movement in the researches, as well as the lack of comprehensive knowledge bases, especially in the domestic stock market (Iran), the necessity of creating and using a comprehensive ontology for Persian financial markets seems to be confirmed. As mentioned in the introduction, with the aim of using ontology in more accurate forecasting of the domestic stock market, previous Persian ontology of the stock market and financial markets, which includes 565 concepts, 496 hierarchical relationships and 137 non-hierarchical relationships and 937 examples, has been developed by the authors of this research [2]. In the present research, in order to create comprehensiveness and standardization of the created ontology, an attempt has been made to strengthen and enrich the ontology using other global ontologies related to the subject of the stock market, and the necessary resources to do this have been reviewed in section 2.2.
2.2 Introduction of knowledge resources related to financial markets
In the scientific space of financial fields, many knowledge bases have been introduced and used. In general, there is no definitive or universally agreed upon knowledge base that all stock market analyzes refer to. The choice of knowledge base depends on various factors, including the specific research objectives, the expertise of the domain analysts, and the availability of relevant resources.[10]
However, when performing stock market analysis, analysts often rely on a combination of different sources of knowledge, including:
1. Financial databases: These are comprehensive repositories of financial data, such as stock prices, historical performance, financial statements, and company profiles. Examples include Bloomberg, Thomson Reuters and Yahoo Finance. While these databases are not fully knowledge bases, they provide valuable information for analysis.
2. Economic Indicators: Analysts often consider macroeconomic indicators such as GDP growth, interest rates, inflation rates, and employment data. These indicators provide insights into the overall health of the economy and can influence stock market trends.
3. Financial Research Reports: Analysts rely on research reports from financial institutions, investment banks and independent research firms. These reports provide in-depth analysis, forecasts and recommendations on specific stocks or sectors.
4. Academic Research and Publications: Researchers may refer to academic articles, journals, and publications that focus on stock market analysis, financial modelling, and related topics. These resources contribute to the existing body of knowledge and provide insights into various quantitative and qualitative analysis techniques.
5. Regulatory information: Analysts often refer to regulatory filings such as annual reports, quarterly reports, and proxy statements. These filings contain important information about the financial performance, risks, and governance of publicly traded companies.
6. Industry-Specific News and Reports: Analysts may follow industry-specific sources of news, market reports, and expert opinions to learn about developments and trends in specific sectors or industries.
The important point is that the field of stock market analysis is dynamic and new sources and knowledge bases emerge over time. Analysts constantly review and evaluate various sources of information to enhance their understanding and make informed investment decisions. The mentioned items are the input of the created anthologies in this field, the most important of which are mentioned below.
1. Financial Industry Business Ontology (FIBO): It is a comprehensive ontology that covers a wide range of financial concepts including stocks, bonds, currencies and commodities. This knowledge base provides a structured representation of financial data and can be used to facilitate the integration of financial information in different systems. [11]
2. Financial Instrument Global Identifier (FIGI) Ontology: The FIGI ontology is designed to provide a standard identifier for financial instruments such as stocks, bonds and options. It can be used to link different sources of financial data and provide a consistent method to identify financial instruments in different systems. [12]
3. Stock Market Ontology (SMO): A domain-specific ontology that provides a structured representation of stock market data. It covers concepts such as stock price, trading volume and company finances. It can be used to facilitate the integration of stock market data in different systems and provide a basis for stock price prediction models [6].
In the following, these ontologies will be described in order to use them in enriching the basic ontology.
2.2.1 Financial Industry Business Ontology (FIBO)
This knowledge base is a standard and comprehensive ontology that aims to define concepts, relationships and laws in the field of the financial industry. FIBO provides a common language and framework for the presentation and exchange of financial information across different systems and organizations and covers a wide range of financial areas, including securities, derivatives, banking, loans and corporate actions. This ontology is designed to depict complex relationships between financial entities such as issuers, investors, intermediaries and regulatory bodies, and defines concepts such as financial instruments, parties involved in financial transactions, market data, pricing and risk management. and incorporates industry best practices, standards and regulations to ensure consistency and accuracy in the presentation of financial data. [11]
FIBO can be used in a variety of applications in the financial industry, including regulatory compliance, risk management, data integration, and financial analytics. By using FIBO, financial institutions can increase data interoperability, reduce data integration costs, and improve overall data quality and consistency.
Overall, FIBO serves as a basic framework for modeling and presenting financial knowledge, enabling efficient data management and facilitating interoperability within the financial industry and offers a rich set of relationships and rules (including ownership, classification, aggregation, dependency, law and time) that will enable detailed visualization and analysis of financial data and support various use cases in the financial industry such as risk management, regulatory compliance and financial reporting. [11]
An example of the relationships in this ontology is given in Table 1.
Table 1 Example of FIBO ontology relationships
The name of the relationship
|
Example
|
Ownership relation
|
Company A owns 50% of the shares of Company X.
|
Classification relation
|
Financial instrument Z is classified as a bond in the fixed income asset class.
|
Aggregation relation
|
Portfolio P1 aggregates the performance of individual financial accounts A1, A2 and A3.
|
Dependency relation
|
The valuation of contract A depends on the market data M1 and M2.
|
Rule-based relation
|
Credit risk limit rule: The exposure to a counterparty must not exceed a specified credit limit.
|
Temporal relation
|
Historical price of Stock A on a specific date was $100 per share.
|
2.2.2 Financial Instrument Global Identifier (FIGI) Ontology
The Financial Instrument Global Identifier (FIGI) Ontology is a standardized framework that provides a consistent and unique identification system for financial instruments. It is designed to facilitate the accurate and efficient identification and classification of financial instruments across various systems and platforms. [12]
The FIGI Ontology defines a set of concepts, relationships, and attributes to represent the characteristics and metadata associated with financial instruments. It enables the integration and interoperability of financial instrument data by providing a common language for identifying and categorizing instruments.
The ontology includes concepts such as FIGI codes, security types, asset classes, exchange codes, and issuer information. It captures the relationships between these concepts, allowing for the classification, grouping, and querying of financial instruments based on their attributes.
The FIGI Ontology supports various use cases in the financial industry, including trading, portfolio management, risk analysis, and regulatory reporting. By utilizing FIGI codes and the associated ontology, market participants can achieve accurate and standardized identification of financial instruments, leading to improved data quality, reduced operational risk, and enhanced data integration capabilities.
Overall, the FIGI Ontology provides a robust and standardized framework for identifying, classifying, and organizing financial instruments. It serves as a foundation for efficient data management, enabling seamless integration and interoperability of financial instrument data across different systems and platforms. The relations and rules in the FIGI ontology (including Is-a, Has attribute, reference, is-denominated-in, Has exchange and rule-based) enable the precise identification, classification and organization of financial instruments. By using these relationships and rules, market participants can ensure stability, interoperability and improve data quality in identifying financial instruments in different systems and frameworks. [12]
An example of the relationships in this ontology is given in Table 2.
Table 2 Example of FIGI ontology relationships
The name of the relationship
|
Example
|
|
Is-a relation
|
Equity ABC is an equity instrument.
|
References relation
|
Equity ABC references Issuer Company X.
|
Is-denominated-in relation
|
Bond XYZ is denominated in EUR.
|
Has-exchange relation
|
Equity ABC is traded on Exchange A.
|
Rule-based relation
|
FIGI code rule: A FIGI code must follow the format AAAA99999999.
|
2.2.3 Stock Market Ontology (SMO)
The Stock Market Ontology is a knowledge representation framework that aims to capture the concepts, relationships, and rules related to stock markets and their associated entities. It provides a standardized structure for modeling and analyzing stock market data, enabling better understanding, integration, and utilization of information in the domain of stock trading and investment. [Zdraveski et al., 2017]
The ontology encompasses various key components of stock markets, including stocks, exchanges, trading activities, market data, market participants, and regulatory aspects. It defines concepts such as stock symbols, stock prices, trading volumes, market indices, trading orders, and investor profiles.
The Stock Market Ontology captures the relationships between these concepts, allowing for the representation of stock market dynamics and interactions. It represents relationships such as stock ownership, trading relationships between buyers and sellers, market data dependencies, and regulatory compliance rules.
By utilizing the Stock Market Ontology, stakeholders in the stock market domain, including investors, traders, analysts, and regulators, can gain insights into market behavior, perform data analysis, develop trading strategies, and monitor compliance. It provides a standardized framework that promotes interoperability and data integration across different stock market systems and platforms.
Overall, the Stock Market Ontology serves as a foundation for representing and analyzing stock market-related information. It enables a better understanding of stock market dynamics, facilitates data integration, and supports various applications in stock trading, investment analysis, and regulatory compliance. The relationships and rules in SMO (including ownership, trading, market data, regulatory, classification, financial performance and time-based relationships) enable the display, analysis and integration of stock market data. Using these relationships and rules, stock market stakeholders can gain insights, perform data analysis, develop investment strategies, and ensure compliance with regulatory requirements. [Zdraveski et al., 2017]
An example of the relationships in this ontology is given in Table 3.
Table 3 Example of SMO ontology relationships
The name of the relationship
|
Example
|
Ownership relation
|
Institutional investor ABC holds a 10% stake in Company XYZ.
|
Trading relations
|
Institutional investor XYZ executes a trade on Exchange A.
|
Market data relations
|
Stock price of Company ABC influences the performance of Market Index X.
|
Regulatory relations
|
Trades above a certain threshold must be reported to the regulatory authority.
|
Classification relations
|
Stock XYZ is categorized under the Technology sector.
|
Financial performance relations
|
Stock XYZ's price-to-earnings ratio is 15.
|
Time-based relations
|
Moving averages provide insights into short-term and long-term stock price movements.
|