Introduction
Graph analytics are gradually becoming the face of airline revolutions as a solution to capture the complexity of airport networks. Graphs have been among computer science’s most accurate and popular concepts for quite some time. Technology has enhanced the attractiveness and efficiency of airport services, where all the transportation operations are associated with intermediate routes. Determining a quick, short, and cheap path for transportation creates a continuous indicator of the airline’s operational structure. The airport connectivity forms a network that allows passengers to transfer without check-in. Therefore, the networks involve extensive data, massive calculations, and estimations to identify the most appropriate route.
Background Information
Initially, the graph theory is obtained from the 18th-century mathematician Leonhard Euler (1707-1783). During his era, it was possible to walk around cities, but there was a need to know the crossing point of paths to figure out the shortest distance. The concepts used to find the intersection points resulted in modern graph representations. Tested over time, Euler’s idea has yielded the graph theory, which grew into a branch of mathematics and was later incorporated into computer science for data visualization (Jitendra et al., 2020, p. 80). Using machine learning, graph analytics allows computers to detect cycles, formulate structured data, and discover the connection between data visualizations (Valput, 2019).
Critical Review of Literature
Introduction to Graph Analytics for Big Data
The application of graphs has been a forthcoming process because the aviation industry must continue planning for improvement. Simplified data analysis and pattern identification involve structured assessment and strategic alignment of achievable decisions through graph analytics. As such, a sequence of words cannot adequately describe the start of airport data graphs or offer clarity in traversing it (Moertini & Adithia, 2021, p. 46). Experts have learned that airlines significantly impact economic and scientific perspectives (Saleena et al., 2018, p. 710). In response to this significance, several frameworks have been designed to deal with big numbers, given the availability of data and the complex agent interactions (Moertini & Adithia, 2021, p. 46). Using graph theory and principal component analysis, one can efficiently study and discover the most popular flights and other aspects of the aviation industry.
The complex network theory was the first successful application of airline traffic analysis as it accounts for multiple factors (Guo et al., 2019, p. 10). For instance, one can analyze the shortest path to identify a complex speed increment that reduces delays and airline congestion. The complexity theory classifies analysis in terms of lower order and high order communities (Yang & Le, 2021, p. 9378). With advanced technologies, the aviation industry has applied computerization and automation of mathematical tools. Graph analytic approaches aim to consider the customer’s choice and uncover the common statistical patterns that infer latent knowledge (Mehta et al., 2022, p. 700). Graphs are significantly prevalent in many domains with their design optimizing workflow (Dave et al., 2016, p. 5). The ability to read and understand physical operations stimulates realistic conditions to operate.
Data Generation, Integration, Management, and Storage
Data extraction and analysis have been in the face of machine learning and graph analytics for some time. Research shows that graph analytics is a popular real-life dataset, gathering information from available flights and airport databases (Mehta et al., 2022, p. 700). Critical reviews show that machine learning has been a unique application of software in graph representation but is not tied to any data formats. However, generating information within data suffers from poor model performances and has limited features represented by predefined formulas (Yuan & Li, 2021, p. 70). In ideal conditions, the aviation industry has suffered from computational constraints and time-consuming data preparation procedures.
The high demand for air traffic opens new windows to examining limited data sources with real-time accuracy. As a result, applying machine learning techniques meets the excessive requirements to run data analytics effectively, primarily when tasked with massive data available in the aviation industry (Valput, 2019). In the airline industry, data grows increasingly due to the development of airport networks (Mehta et al., 2022, p. 700). Technological approaches have been implemented to maximize computation power, including characteristic spatial distribution of the delays across airport networks (Zhou et al., 2022). Graphs are built on changes as data increases, so the replication complex demands are continuously met to detect emerging behaviors from travelers. However, as time goes by, the method has limited engineering applications as it is only effective when established on precise trajectory data.
Mathematical Tuning
In today’s aviation industry, the signal phase in a traffic movement is characterized by stages. The stages represent scenarios where airline vessels’ conflict probability integrates vertical and horizontal arrival chances (Ma et al., 2022, p. 11). These probabilities are validated based on the performance of the proposed models used to measure and collect data from the intersection point (Wong et al., 2020, p. 102068). In short, graphs are generated from conflict simulations with the ability to acquire real-time data where the risks of the conflict are signalized as coefficients at the intersection point. These algorithms are then automated and used to generate graphs that could take any communicative format.
Analysis of graph signals ensures that the delays and dynamics narrow down to derive insights from the technical intelligence given by big data. In this light, graph analytics have improved and advanced their compatibility with machine learning to meet the aviation industry’s primary traffic management goals (Zhu et al., 2022, p. 105). Computer scientists have designed sophisticated algorithms that collect aviation data and match it to the most recent formulas before generating graphs. With the most recent advancements, data presented in graphical formats are undoubtedly accurate because it uses non-parametric approaches to understand transportation delays (Saleena et al., 2018, p. 710). It is projected that graphs and artificial intelligence can collect data from waveforms of moving planes, people, or vehicles around the airport (Guo et al., 2019, p. 10). These waves would then convert into real-time data to produce graphs with competitive results.
In the aviation industry, a sorting comparison network (SCN) model to identify the open chances of growth within the complex airline networks. The new phase of intelligent airports has been invented where integrated systems are used to deliver solutions and ensure an efficient aviation ecosystem. Experts have used innovative approaches to understand the business environment around airports and create a sharable viewpoint for the management (Narongou & Sun, 2021, p. 220). In many cases, business analysts study consumer disparities and develop proposals for the new functions that satisfy customers the most. The data-driven recommendations have increased the sensitivity and autonomy of businesses using artificial intelligence. The analysis creates room for advanced ICT and other bits of intelligence, allowing people to access a system through internet-connected devices.
Airline Instrumental Panel
The spanning and signaling process eases data simulation and enhances accurate prediction of the general traffic situation between the two airports. Incidentally, airline data analysis software suffers from data overloads, gets slow, and fails to coincide with the preferred timelines for a traveler to move from one port to another (Mehta et al., 2022, p. 700). The instrumental panels in the weighted network systems represent all the numerical properties associated with passengers and the capacity an airport can hold. As the human brain struggles with such errors, computers find it hard to store the large graph in a tensor or feed it to an existing algorithm, especially when it involves combined analytical methods. Given the room for failure when graph analytics are built on real-time trajectory data, more research has been shown to project signals when comprehensive data is unavailable.
Despite various methods to predict and evaluate traffic, new approaches have emerged to explain aviation network construction. Ideally, the data available from commercial and academic databases have been used to structure transportation statistics and geo targets that can be used to develop a whole network of airports (Mehta et al., 2022, p. 700). The discovery of more accessible means to collect data from waves simplifies the identification of ports to associate with based on the number of flights a network can accommodate. Given that graph analytics can be configured with different forms of data, it can be used to represent the aviation system entities, relationships, strengths, and weaknesses.
Graph analytics is majorly known as the parenting studies that resulted in the inter-channel attention mechanism because machine learning can decode and use the most relevant data. A correlation between the massive data passed through graphs optimizes the management’s investments in operation improvement activities (Mehta et al., 2022, p. 700). Moreover, the details are used for financial planning, and economic diversity as the innovative airport management understands its customer base, needs, and primary practices.
Airline Industry Forecasting
As the fastest means of transport, airlines projected in graphs resolved the mismatch between the traffic demand and constrained capacity of resources. Among the major complaints that have faced air logistics is congestion problems. Despite constant improvements, these problems are expected to be prolonged. Graphs ensure an intense collaboration between industrial practitioners and academics to add new value and knowledge to solve air traffic issues (Li et al., 2021, b, p. 560). The analysis of air transport allows researchers to understand passengers’ trends and employment opportunities to ensure all aspects of the airline network run effectively.
As the volume of data grows, the world shifts to extensive data analysis as a response to derive desired business values. Graph analytics visualizes real-world problems and narrows the investigation to a given industry (Jitendra et al., 2020, p. 80). Economists also study the minimum cost path a plane can take from one destination to another. Moreover, analysts focus on the ever-changing nature of the market and industry demands to break down complex information across structured and unstructured data blended from different applications. In this line of analysis, literature shows that research examines the optimal trips and the conditions under which they happened (Gopalakrishnan & Balakrishnan, 2021, p. 400). With these details, the airline companies add on the number of trips falling under these categories. Graph analytics are also used to evaluate the scalability and survivability of an applied strategy, especially when it involves significant changes in the entire operating model of the company.
Air transport is affected by multiple factors that control traffic, for example, the economic situation after the Covid-19 pandemic. Due to the need for improved planning, pandemics create significant difficulties in identifying fuel costs, government regulations, and analyzing tension-induced passenger flocking in the airport. The heavy air traffic is considered a constraint to primary operations, given that half of the world’s population is carried by planes on a yearly basis (Li et al., 202, a, p. 15). As such, there has been an increasing need to accurately determine and characterize the objective dynamic process of delayed propagation between ports.
Graph analytics have empowered artificial intelligence by ensuring that experts use uniform approaches to aid in rational reasoning. Using the graph network analysis creates new opportunities to discover as many airport connections as one may need to get tickets slotted for a desired time or date (Zhu et al., 2022, p. 105). Besides, graph analytics help in the combined generalization of trends and can be used to prioritize flights. The ability of artificial intelligence to learn from graphs ensures that they acquire human-like capabilities in managing air traffic and deriving conclusions about the stability of an airline system.
Literature shows that the academic literature on graph analytics in the aviation industry involves statistical, operation management, and econometric models (Saleena et al., 2018, p. 710). Through technology and computer science, conceptualization and development of hard-to-write graph generation software have been implemented. The automation of graphs has become a new research component in aviation software engineering guided by complex test stubs to resolve the open concept constraints in air traffic management (Semeráth et al., 2021, p. 1720). In the social network analysis, experts use auto-generated graphs to understand the relationship between nodes and edges in airport networks.
Graphs provide the most reliable and convenient visual feedback where traffic managers and travelers can predict and combine the entire neural network. As such, the networks between airports obtain and share information amongst the interconnected parties describing the roads and traffic state (Qu et al., 2022). Such data is later fed into graph analytics to help monitor the situation and grade the degree of expected effects of traffic congestion (Zhang et al., 2019). The classification and experimentation of the traffic prediction outcomes using airport-generated data often require formula weighting calculations. As a result, a combination of distance and relative speeds can be estimated in ways that resolve short-term traffic situations.
Literature has studied the signalized intersection conflict probability model as precise anticipation of straight-left traffic flow. The model uses the heat map to evaluate traffic safety and prebuild signalized intersections on urban roads (Li et al., 2021, a, p. 15). This approach is so sophisticated that it analyzes the impact of vehicle movements on the characteristic traffic conflict viewed at the signalized intersection points (Guo et al., 2019, p. 10). The analysis is brought to effect by incorporating vehicle movement trajectory characteristics in graph analytics.
At the airports, traffic analysis is done to ensure signalized intersection safety points have complete control of the traffic flow, so accidents and crushes do not happen regularly. The existing literature focus is the statistical prediction that can be read from different graphs and the spatial-temporal characteristics of traffic conflicts (Zhu et al., 2022, p. 105). In most cases, the statistical methods predict potential matches hoping to warn early in advance (Gopalakrishnan & Balakrishnan, 2021, p. 400). Most experts explored the spatial-temporal distribution maps as an adequate evaluation of traffic and airline safety.
Most scholars are evaluating how graphs can identify critical industrial problems and interests. A new window to identify and assess opportunities is slowly emerging. Recently, analytics has been considered a vital part of understanding and improving the profitability margins of many businesses (Dai et al., 2020, p. 3080). In this window, the airline industry is not an exception. With different market graph analytics modes, descriptive, predictive, and prescriptive data analysis ensures airports understand the awaiting opportunities.
Advanced Booking
Graphs collect and present temporal, long-term, and repetitive patterns in air traffic, allowing the booking agents to understand the characteristics of multi-distribution traffic within airport-generated data. Ideally, the analysis captures all activities within the airport and analyzes them to develop better process cycles that enhance customer experience (Bunchongchit & Wattanacharoensil, 2021, p. 100688). For instance, airports and booking agents have learned the demographic configuration of the patients traveling with a piece of luggage and invested in pick-up services at a fee. The technologically connected aviation ecosystem allows travelers to book a flight using mobile payments.
The graph theory has continuously been applied to model and describes all the experiences travelers undergo and detail the shortages or available possible improvements. The booking agencies use charts to predict airport traffic at micro, meso, and macro levels (Zhu et al., 2022, p. 105). For instance, the analysis of delays caused by many travelers indicates the number of attendants to resolve the challenge (Bunchongchit & Wattanacharoensil, 2021, p. 100688). Similarly, a different study exploring the congestion of the airport waiting area creates room for new constructions and the advancement of physical infrastructure. Therefore, research and graph analytics centered on the connected components offer a variety of choices that can be considered improvement points.
Conclusion
Current studies have explored the evolution and development of graph theory and principal uses such as data capture. In these descriptions, it has become clear that graph analytics develop rapid research using performance indicators (Jitendra et al., 2020, p. 80). Similarly, graph analytics has made it easy to project outcomes from certain business operations and all the problems that might accrue. It can be concluded that they give a solid solution to the traffic conditions, airline problems, and a forecast of the viable business options. These findings are informative, but there is still room for more innovations using this set of analytics.
Future Applications and Recommendations
In the future, researchers should take advantage of the capabilities to use graphs and discover the relationship between entities faster and use them to learn more about the aviation industry. For instance, graph analytics can examine the relationship between delays, low fares, currencies, and fuel (Qu et al., 2022). This notion opens a new window to explore how airports recover their path lines after the Covid-19 pandemic. Additionally, the analytics would help explain the competition risks or entry into the market using data from the factors that influence flight operation. Given that air cargo transportation is the most vulnerable part, graph analytics should be used to predict the occurrence of illegal or hazardous shipments. Researchers should develop a clearer picture of where hijacking cases occur most and account for minor but possible accidents for safety and precaution.
It is worth noting that graphs analytics have broken down the struggles like waiting time experienced by travelers and laid down the layout of the airport surroundings to smoothen operations. Besides, the calculated data that allows booking agents to grade airports based on the contained traffic creates room for segmentation (Dai et al., 2020, p. 3080). Most airports can invest in the taxi industry because most customers prefer being escorted after a flight or facilitating pick-ups for the next flight.
The analytics are also commendable for the benefits of digitization in airport facilities. Indeed, the management understands the investment capacities and all the risks expected through a complex future forecast. Within proximity, airports can develop mergers with hotels and accommodation facilities to reduce client waiting time (Bunchongchit & Wattanacharoensil, 2021, p. 100688). The customizability of these analytics makes it easy to use them for internal reviews, especially when there is an eternal need to define a solution.
References
Bunchongchit, K., & Wattanacharoensil, W. (2021). Data analytics of Skytrax’s airport review and ratings: Views of airport quality by passengers types. Research in Transportation Business & Management, 41, 100688.
Dai, R., Xu, S., Gu, Q., Ji, C., & Liu, K. (2020, August). Hybrid spatio-temporal graph convolutional network: Improving traffic prediction with navigation data. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (pp. 3074-3082).
Dave, A., Jindal, A., Li, L.E., Xin, R., Gonzalez, J. and Zaharia, M., 2016, June. Graphframes: an integrated API for mixing graph and relational queries. In Proceedings of the fourth international workshop on graph data management experiences and systems (pp. 1-8).
Gopalakrishnan, K., & Balakrishnan, H. (2021). Control and optimization of air traffic networks. Annual Review of Control, Robotics, and Autonomous Systems, 4(1), 397-424.
Guo, W., Toader, B., Feier, R., Mosquera, G., Ying, F., Oh, S. W., Williams, P. M., & Krupp, A. (2019). Global air transport complex network: multi-scale analysis. SN Applied Sciences, 1(7), 1-14.
Jitendra, M. S., Amiripalli, S. S., Kollu, V. V. R., Chowdary, P. R., & Rao, R. V. (2020). Analysis of airline connectivity system using graph theory. International Journal of Control and Automation, 13(4), 77-84.
Li, D., Wu, J., & Peng, D. (2021, a). Online traffic accident spatial-temporal post-impact prediction model on highways based on spiking neural networks. Journal of advanced transportation, 2021, 1-20.
Li, M. Z., Gopalakrishnan, K., Pantoja, K., & Balakrishnan, H. (2021, b). Graph signal processing techniques for analyzing aviation disruptions. Transportation Science, 55(3), 553-573.
Ma, Y., Zhang, Z., & Wu, J. (2022). Conflict probability prediction and safety assessment of straight-left traffic flow at signalized intersections. Journal of Advanced Transportation, 2022, 1-14
Mehta, N., Ruparelia, A., Verma, J. P., & Khinchi, M. K. (2022). Graph-based data analysis in big data computing environment: An investigation of flight network datasets. In Proceedings of 2nd International Conference on Artificial Intelligence: Advances and Applications (pp. 699-710). Springer, Singapore.
Moertini, V. S., & Adithia, M. T. (2021). Uncovering active communities from directed graphs on distributed spark frameworks, case study: Twitter data. Big Data and Cognitive Computing, 5(4), 46.
Narongou, D., & Sun, Z. (2021). Big data analytics for smart airport management. In Intelligent Analytics with Advanced Multi-Industry Applications (pp. 209-231). IGI Global.
Qu, H., Guo, J., & Jiang, Y. (2022). Research on recommendation algorithm of joint light graph convolution network and DropEdge. Journal of Advanced Transportation, 2022.
Saleena, P., Swetha, P. K., & Radha, D. (2018). Analysis and visualization of airport network to strengthen the economy. International Journal of Engineering & Technology, 7(2), 708-713.
Semeráth, O., Babikian, A. A., Chen, B., Li, C., Marussy, K., Szárnyas, G., & Varró, D. (2021). Automated generation of consistent, diverse and structurally realistic graph models. Software and Systems Modeling, 20(5), 1713-1734.
Valput, D. (2019). Machine learning with graphs: The next big thing? Web.
Wong, A., Tan, S., Chandramouleeswaran, K. R., & Tran, H. T. (2020). Data-driven analysis of resilience in airline networks. Transportation Research Part E: Logistics and Transportation Review, 143, 102068.
Yang, H., & Le, M. (2021). High-Order Community Detection in the Air Transport Industry: A Comparative Analysis among 10 Major International Airlines. Applied Sciences, 11(20), 9378.
Yuan, H., & Li, G. (2021). A survey of traffic prediction: From spatio-temporal data to intelligent transportation. Data Science and Engineering, 6(1), 63-85.
Zhang, W., Yu, Y., Qi, Y., Shu, F., & Wang, Y. (2019). Short-term traffic flow prediction based on spatio-temporal analysis and CNN deep learning. Transportmetrica A: Transport Science, 15(2), 1688-1711.
Zhou, F., Jiang, G., Lu, Z., & Wang, Q. (2022). Evaluation and analysis of the impact of airport delays. Scientific Programming, 2022.
Zhu, X., Lin, Y., He, Y., Tsui, K. L., Chan, P. W., & Li, L. (2022). Short-term nationwide airport throughput prediction with graph attention recurrent neural network. Frontiers in Artificial Intelligence, 105.