Data-Driven Decision Making in Digital Media Management

 

Deepa Dixit ¹, Dr Priyanka Srivastava ², Dr. Swapnali Amol Kulkarni ³, Dr. Raju ⁴

 

¹ SIES School of Business Studies, Navi Mumbai, India

² Associate Professor, MBA - Department of Marketing, Indira Institute of Management Pune, India

³ Associate Professor, School of Business, Indira University, Pune, India

⁴ Assistant Professor, Department of Computer Science and Engineering (AIML), Noida Institute of Engineering and Technology, Greater Noida, Uttar Pradesh, India

 

1. INTRODUCTION

In the modern world of digital transformation, the ability to make decisions based on data (DDM) is now the key to proper management in any industry, and digital media are one of the most active participants Bange and Lorenz (2022). The emergence of digital platforms and the user-generated content boom coupled with the introduction of real-time analytics has transformed the way that media organizations conceptualize, produce, distribute, and assess content. In this changing ecosystem, the decision-making process is less driven by intuition or creative instinct, but more and more is informed by empirical knowledge based on huge amounts of well-organized and unstructured data Chen et al.  (2023). As the audience continues to become more fragmented across systems and platforms, data analytics has become necessary to understand user behavior, to enhance engagement strategies, and to enhance monetization models. The current state of digital media management is in the quandary of technology, creativity, and analytics. Big data analytics, machine learning (ML), and artificial intelligence (AI) help media companies, including streaming platforms (Netflix, Spotify), social media giants (Meta, Tik Tok), and others, to extract actionable insights on the basis of the audience interactions Li et al. (2023). Such insights can empower managers to make a prudent decision in matters of personalization of contents, recommendation systems, advertisement schemes and effectiveness of functions. The predictive decisions are also facilitated by data models due to the ability of the algorithm to project the desires of the audience, the capacity to allocate resources in a beautiful manner and the ability of making the most of the investment (ROI). The good examples are the content recommendation system of Netflix and the adaptive algorithms of YouTube, which can be seen as a good example of how a data-driven approach can be used to keep viewers and make them happier Wang et al.  (2024).

Figure 1

Figure 1 Evolution of Data-Driven Decision-Making in Digital Media

 

The resulting explosion of sources of data is closely linked with Change Digital media ecosystems. The likes, shares, and clicks of all online interactions; the time watched on video leave a footprint in the form of data as presented in Figure 1. Analytics pipelines which operate on a cloud computing platform, data lakes and business intelligence tools are used as tools to summarize this data into valuable patterns Zhang et al. (2024). The media managers interpret these trends in a manner that would enable them to simplify their campaign mode, measure the performance metrics and predict future trends. DDDM, in turn, transforms the management of the digital media into the cycle and repetitive process of measurement, analysis, optimization and feedback that can unite business decisions with evidence-based knowledge Chen et al.  (2024).

 

2. Background History

The latter has been significantly driven by the increase of data accessibility and improvement of computational analytics, which has transformed the digital media industry in the last decade Vaidya et al. (2025). Historically, the qualitative measures editorial intuition, audience surveys and focus groups that offer scanty information about the real-time audience behavior were used to influence media decision-making. As digital platforms and social media emerged, the old paradigm was no longer suitable to consider the scope and intricacy of the contemporary content ecosystems Gao et al. (2017), Lee et al. (2019) The change in technology also led to the increase in the detail and richness of the data available. The combination of the Big Data structures, machine learning techniques, and the cloud computing system has changed the ways in which the media organizations cognize the audience behavior Bueno et al. (2021). Media companies started using real-time analytics to divide audiences, customize content, and maximize the advertising process. This development brought a paradigm shift in the reactive decision-making to proactive decision-making- are predictive insights and not just historical data that help in business strategies Fu et al. (2020).

 

3. Research Design and Method

The design allows the triangulation to increase the validity and reliability of the findings since it covers both the empirical patterns and the interpretative information of practitioners in the industry. The general methodological approach is a deductive reasoning model where the theoretical background starts with the literature on decision theory, data analytics and literature on digital media management as the theoretical basis and is followed by the empirical validation of the results using data collection and analysis. The study takes the form of exploratory-descriptive design. The exploratory dimension is aimed at exploring new trends, structures and technologies in favor of DDDM in digital media, whereas the descriptive dimension is designed to quantify the extent and effectiveness of the practices in the industry. The research follows a series of three steps:

·        Conceptual Modeling: The development of a theoretical framework that connects decision theory, analytics maturity and strategic delivery in media management.

·        Interpretive Evaluation: The qualitative interviews and case analysis as a way of interpreting results in real-world managerial situations.

It is a framework that has guaranteed the quantitative data creates measurable relationships with the qualitative data describing contextual subtleties and managerial implications.

Step-1] Data Collection Strategy

The quantitative analysis sampling size is about 150-200 respondents which is adequate to be used in statistical inferential testing. The interviews with 10-15 industry professionals complement the quantitative data because they provide depth and richness of qualitative data.

Step-2] Research Variables and Constructs

The research conceptualizes a number of constructs based on decision theory and models of media management:

·        Independent Variables: Data Quality, Analytics Maturity, Decision Support Tools and Organizational Culture.

·        Dependent Variables: Accuracy of Decision, Strategic Agility, and Strategic Audience Performance.

·        Moderating Variable: Data Interpretation Managerial Experience.

The statistical models and structural equation modeling (SEM) are used to analyze the relationships among the variables later.

Step-3] Analytical Methods

The quantitative data are examined with the help of the descriptive statistics, correlation analysis, and the Partial Least Squares Structural Equation Modeling (PLS-SEM) to discover the causal relationships between the data-driven capabilities and the outcomes of managerial performance. Predictive regression and clustering machine learning techniques are also included in the classification of decision patterns among respondents.

Interpreted data are analyzed qualitatively by using interviews as a source of data, and thematic analysis (NVivo or the same software) is applied to determine recurring themes according to the processes of decision making, data ethics, and the perceived obstacles in their implementation. The combination of these analyses provides the possibility to cross-verify numerical patterns and managerial experiences, which increases the reliability of result.

 

4. DECAS Framework

The DECAS framework (Decision-making process, dEcision maker, deCision, dAta, and analyticS) serves as an all-encompassing theoretical framework of crediting the role of data analytics in managerial decision-making. It is both a human and technical blend of the decision system acknowledging that good data-driven decision-making (DDDM) is not just about data availability or availability of tools but primarily about how knowledge is decoded and acted upon. The DECAS framework can be used as a perfect template within the framework of the digital media management to interpret how companies like Netflix implement analytics at various levels of decision-making.

Figure 2

Figure 2 Decision-Theoretic Framework for Data-Driven Decision-Making

 

It connects cognitive, organizational, and technological levels and as such both rational (data-based) and behavioral (human-based) factors are taken into account as is illustrated in Figure 2. Through the integration of these areas, DECAS describes the process of data collection to strategic implementation with a focus on the synergy between human-based judgment and analytical intelligence.

Table 1

Table 1 Components of the DECAS Framework
Component	Definition	Role in DDDM (Digital Media Context)	Netflix Case Illustration
D – Decision-Making Process Lu et al.  (2020)	The structured set of steps by which alternatives are evaluated and selected.	Guides how media organizations define problems, collect data, analyze alternatives, and execute choices.	Netflix follows a cyclical decision model: problem definition → data collection → model analysis → A/B testing → outcome assessment.
E – Decision Maker Reveilhac and Morselli (2020)	Individuals or teams responsible for interpreting analytics and making judgments.	Decision makers blend analytical insights with creative and strategic thinking to optimize media operations.	Cross-functional teams of data scientists, engineers, and creative executives jointly determine which shows to greenlight or promote.
C – Decision	The final action or choice resulting from data analysis and human interpretation.	Reflects organizational priorities such as user engagement, market expansion, or innovation.	Data insights informed the decision to produce House of Cards, aligning data-predicted demand with creative feasibility.
A – Data Nie et al. (2017)	Quantitative and qualitative inputs used to inform and validate choices.	Includes structured (user logs, KPIs) and unstructured (reviews, comments, sentiment) data sources.	Netflix collects real-time behavioral, contextual, and demographic data to refine personalization algorithms.
S – Analytics	Tools and techniques that transform data into actionable insights.	Includes AI/ML models, descriptive dashboards, predictive simulations, and prescriptive optimization systems.	Netflix’s deep learning-based recommendation engines and reinforcement learning models guide personalization and marketing strategies.

 

The Netflix case study confirms the main assumption of the present study: the method of data-driven decision-making can improve the efficiency of the operations of a company and increase the satisfaction of the audience when it is integrated into a systematic, theory-driven framework. The decision theory forms the basis of maximizing uncertain results, whereas DDDM implements it with the help of computational analytics.

 

5. Case Study: Netflix Movies

In order to demonstrate the practical implications of Data-Driven Decision-Making (DDDM) in digital media management, the given section provides a case study of one of the brightest examples of how data analytics can transform content management, audience interest, and the optimization of revenues, Netflix. Netflix is the ideal embodiment of a data-driven culture in which all decisions, including what content to buy and how to show it to customers, are informed by the data analytics and ongoing tracking. The case study shows that the principles of the decision theory, along with the use of big data analytics, inform the strategic and operational decisions, which help Netflix to be the global leader in the streaming industry.

Table 2

Table 2 Data-Driven Decision-Making Metrics and Insights from Netflix Case Study
Category	Data Source / Variable	Analytical Method Applied	Decision / Action Taken	Observed Outcome / Key Metric	Interpretation / Managerial Insight
Audience Behavior Analytics	Viewing history, watch duration, pause/resume events	Time-series clustering and association rules	Personalized recommendation feed for each subscriber	80% of viewing activity originates from recommendations	Personalized content increases engagement and reduces churn
Content Preference Analysis Ninawe et al. (2025)	Genre affinity, actor/creator data, viewer rating trends	Bayesian probability models and collaborative filtering	Greenlit original content such as House of Cards based on predicted audience appeal	25% increase in completion rates for personalized recommendations	Data-guided production reduces financial risk and improves ROI
User Interface Optimization	Clickstream data, heatmaps, navigation sequences	A/B testing and behavioral analytics	UI redesign for thumbnails and previews to improve discoverability	15% improvement in content selection speed and 12% reduction in bounce rate	Real-time experimentation enhances user satisfaction and interface usability
Marketing Strategy	Regional search trends, social sentiment, subscription patterns	Sentiment analysis and predictive regression	Tailored ad campaigns by region and language	18% higher conversion rate in targeted markets	Localized marketing improves cultural resonance and customer acquisition
Technical Performance Monitoring	Playback latency, buffering rate, device-type data	Network analytics and QoS monitoring	Dynamic bitrate streaming and caching strategy updates	22% decrease in buffering-related session drop-offs	Integrating technical metrics into DDDM enhances platform reliability

 

Technical Performance Monitoring Playback latency, buffering rate, data of device-type Network analytics and quality of service monitoring Dynamic Bitrate streaming and caching strategy update 22% fewer sessions dropped off because of buffering issues

The hybrid system of Netflix which is learning based and personalize the recommendations based on more than 1,000 user and content factors. Though Netflix can be considered as a success case, all media organizations are not able to emulate Netflix because of the high cost of investments, complexity of technologies, and restrictions of data access. Moreover, over-dependence on predictive analytics might impair the creative variety in the event that algorithms are too focused on the past.

 

6. Applying the DECAS Framework in the Netflix Case Study

Using the DECAS (Decision-making process, dEcision maker, deCision, dAta, and analyticS) framework in relation to the Netflix case study offers a systematic view of how the company manages to utilise data analytics in the process of making informed, timely, and creative decisions. The DECAS lens can be applied to the operation and strategic model of Netflix as a closed-loop ecosystem, in which human expertise, algorithmic intelligence, and feedback constantly work together to produce high-quality, evidence-based decisions, as illustrated in Figure 3. This section breaks down every DECAS element as it can be seen in the Netflix data-driven setting, showing how the company has made a culture of decision-making excellence analytical.

Figure 3

Figure 3 Data-Driven Decision-Making Framework Applied at Netflix

 

This assessment uses a five-dimensional model commonly applied in digital transformation studies:

1)     Analytical Capability and Adoption of Tools.

2)     Organizational Culture and Data Literacy.

3)     Process integration of decisions.

4)     Governance, Ethics and Compliance.

Each dimension is rated on a five-level maturity scale (1 = Initial, 5 = Optimized).

Step -1] D — Decision-Making Process

The process of the decision making at Netflix is very iterative and evidence based. Since Netflix is based on data, there is a cyclical process of problem identification, data extraction, model development, testing, implementation, and performance feedback as opposed to the traditional media houses that rely on intuitive and market assumptions. Netflix first needs to understand the demand indicators in its huge database - genre usage, completion rates, or regional interests before greenlighting a new series. The insights are modelled with the help of predictive analytics and examined with the help of A/B and multivariate testing. Since a show has been launched, post-decision analytics measure its performance, enhancing later decision-making. The dynamic loop is consistent with the idea of the framework of DECAS that there is no single decision-making episode but rather an ongoing, dynamical process that is made by the feedback.

Step -2] E -- Decision Maker

At Netflix, a hybrid organization exists with a decision maker being a human-AI system. The decision-making power is decentralized instead of being centralized with various teams having the power to take independent action within specified parameters of analysis. The culture of Freedom and Responsibility at Netflix guarantees that the data information is used to empower people to make informed decisions and not to take up their creativity. An example is that although algorithms propose the best content acquisition strategy, creative originality, cultural influence, and artistic value are given the last approving mind.

Step -3] C -- Decision

All decisions at Netflix, be it strategic (content investment), tactical (marketing design), or operational (recommendation ranking), are based on the power of analytics. The decisions made are based on maximization of expected utility and they are based on principles of decision theory within the DECAS model. Netflix uses probabilistic models (score of expected engagement), possible regional popularity (market segmentation), and cost of production (budget constraints) when choosing new shows. This multi-criteria evaluation makes it possible to make risk-optimal decisions. Indicatively, data modeling anticipated that the Korean series Squid Game was going to achieve outstandingly well in various international markets because of the cross-cultural narrative drive, as well as the existing exposure to Korean contents, which confirmed the DECAS dimension of decision in practice.

Step -4] A -- Data

Information is the blood of the Netflix. The company receives and processes more than 3 petabytes of user-generated data daily on devices, platforms, and locally. These data types include:

·        Contextual Data: Type of device, speed of network, use during the day and location.

·        Demographic Data: The age, language preference, and geographical demographics.

Netflix has a high level of data readiness through the DECAS lens, as the company has consistent pipelines of collection, cleaning, integration, and analysis. The Data Mesh architecture of the company enables decentralized ownership of data and quality and interoperability of data among teams. This helps in flawless incorporation of insights into decision workflows, which makes it real-time responsive..

Step -5] S — Analytics

Netflix relies on descriptive, predictive and prescriptive analytics to make sense of the user behavior and maximize results.

·        Descriptive Analytics: Summarizes viewer engagement and identifies consumption trends.

·        Predictive Analytics: Forecasts viewing patterns using AI/ML models, such as collaborative filtering and deep neural networks.

·        Prescriptive Analytics: Recommends actions (e.g., which content to highlight, when to release, and how to personalize marketing).

 

Table 3

Table 3 Evaluation of DECAS Application in Netflix
DECAS Element	Netflix Implementation	Effectiveness (1–5)	Key Outcome
Decision-Making Process (D)	Data-driven, iterative, and experiment-based	5 – Highly Effective	Continuous refinement of strategy based on feedback
Decision Maker (E)	Hybrid (AI + human) model with distributed autonomy	5 – Highly Effective	Balances algorithmic precision and creative judgment
Decision (C)	Multi-criteria, evidence-based strategic choices	5 – Highly Effective	High audience satisfaction and financial efficiency
Data (A)	Real-time, multi-source, and high-volume data pipelines	5 – Highly Effective	Enables personalization and rapid insight generation
Analytics (S)	Advanced ML, AI, and reinforcement learning systems	5 – Highly Effective	Drives adaptive recommendations and optimized engagement

 

The S-component of DECAS is reflected in the recommendation system that provides the content viewed by 80% of the audience.

 

7. Results and Analysis

Results and analysis part explains findings of the case study by Netflix in terms of Data-Driven Decision-Making (DDDM) and DECAS framework. There will be a combination of quantitative (performance measures, engagement rates, readiness scores, etc.) and qualitative data (e.g., managerial views, cultural adjustment, and ethical aspects). This holism focuses on the way data and analytics and human decision-making interact within the Netflix model of operation to create a measurable output in efficiency, engagement with users, and agility. The quantitative analysis is concerned with the impact of the implementation of a systematic DDDM model by Netflix on such key performance indicators (KPIs) as user engagement, retention, recommendation accuracy, and the payback period (ROI). The secondary sources were also used to synthesize data with the performance reports of Netflix and the already published analytics.

Table 4

Table 4 Quantitative Results of Data-Driven Decision-Making at Netflix
Metric	Before DDDM Integration	After Advanced DDDM Integration	% Improvement	Interpretation
Average Viewing Time per User (hrs/day)	1.8	3.2	77.70%	Personalized recommendations increased session duration.
Churn Rate (%)	7.5	4.1	–45.3%	Predictive churn models improved retention strategies.
Content Completion Rate (%)	62	85	37.10%	Data-driven content curation aligned content with audience preferences.
Recommendation Accuracy (F1 Score)	0.68	0.89	30.90%	Hybrid collaborative filtering enhanced personalization.
Revenue Growth (Billion USD, annual)	20.2	33.7	66.80%	Analytics-driven programming decisions contributed to higher subscriber base.
Customer Satisfaction Index (1–100)	74	92	24.30%	Data-informed UX and service optimization improved satisfaction levels.

 

The data indicate that it has an obvious upward trend in all the indicators of performance after the systematic implementation of DDDM and advanced analytics.

Figure 4

Figure 4 Correlation Between DECAS Readiness and Organizational Outcomes

 

This confirms the hypothesis that a structured analytical system (DECAS) is associated with a positive relationship with better decision efficiency and organizational performance. The third visualization is a scatter plot with a regression trend line that illustrates the correlation between the DECAS readiness of Netflix (X-axis) and the organizational performance outcomes (Y-axis), which is a combination of key indicators as shown in Figure 4, ROI, and satisfaction. The points illustrate a data point in various time periods or areas of operation. The plot shows the positive linear trend is high, which means, the higher the index of DECAS readiness, the higher the results of the organization. The calculated correlation coefficient (r [?] 0.89) indicates an extremely high positive correlation between the DDDM maturity and business performance. The trendline represents a positive movement that, on the one hand, suggests the visual message that there is a strong correlation between data governance, analytics, and decision empowerment as it promotes corporate efficiency and user engagement.

Table 5

Table 5 DECAS Framework Results for Netflix
DECAS Dimension	Key Indicators Observed	Effectiveness Score (1–5)	Supporting Evidence
D – Decision-Making Process	Structured, iterative, feedback-driven decision cycles	5 (Optimized)	Continuous A/B testing, adaptive content scheduling
E – Decision Maker	Cross-functional collaboration between creative teams and data scientists	5 (Optimized)	Distributed decision autonomy with strong data fluency
C – Decision	Strategic decisions backed by predictive analytics	5 (Optimized)	High ROI from content investments (e.g., Squid Game, Wednesday)
A – Data	Scalable, multi-source, real-time data infrastructure	5 (Optimized)	Petabyte-level daily ingestion and structured governance
S – Analytics	Deep learning, reinforcement learning, sentiment analysis	5 (Optimized)	80% of content viewed via recommendation algorithms

 

An application of the DECAS framework generated a list of readiness and performance deliverables that measure the maturity of Netflix in terms of analysis. These results verify that Netflix is at full DDDM maturity. All decisions, both micro-level personalization and macro-level business strategy are based on quantitative and qualitative evidence-making DECAS a valid descriptive and prescriptive model of media decision systems.

Figure 5

Figure 5 DECAS Effectiveness Radar Chart

 

The radar chart graphically represents the maturity and effectiveness of Netflix in the five major elements of the DECAS framework Decision-Making Process (D), Decision Maker (E), Decision (C), Data (A), and Analytics (S). The scales of each axis have 0-5 rising points, which are the levels of preparedness and optimization. The polygon that is filled in the radar chart stretches in all directions up to the maximum level (5/5) to create almost a perfect pentagon. The shape shows that Netflix portrays balanced and optimized competency in each of the dimensions of DECAS as shown in Figure 5. The competitive advantage of the company is maintained by its highly organized decision making process, empowered and knowledgeable decision-makers, high data infrastructure and high quality analytics. The fact that the chart is uniform leads to the conclusion that Netflix is not merely a company that achieved its success in the technical and analytical aspects only but also managed to incorporate human judgment and governance systems into the decision process. Through the radar, therefore, Netflix has the holistic preparedness to make sustainable, intelligent, and adaptive decisions.

 

8. Conclusion

As the paper shows, the adoption of systematic data-driven decision-making (DDDM) that is facilitated by the DECAS framework is the key to the success of Netflix in the field of digital media management. By combining structured decision processes, empowered decision makers, quality data and streamlined analytics, Netflix has turned decision-making into a predictive and reactive intelligence system, rather than an intuitive process. The quantitative data is evidence of the impressive performance increase in major indicators, such as the engagement, reduction of churn, and rise in revenues, and the qualitative one proves that the organization is characterized by a high rate of analytical literacy and innovative collaboration. The application of the DECAS model shows that only data alone is insufficient to create value, the combination of human judgment and computational analytics would create strategic accuracy. Netflix case illustrates the example of full-fledged DDDM ecosystem with ongoing feedback, ethics of data governance, and evidence-based learning as the key tenets of long-term competitiveness.

 

CONFLICT OF INTERESTS

None. 

 

ACKNOWLEDGMENTS

None.

 

REFERENCES

Bange, C., and Lorenz, N. (2022). BARC Data Culture Survey 22: How to Shape the Culture of a Data-Driven Organization (Topical survey). Tableau.

Bueno, R. A., Carrasco, R. A., Porcel, C., Kou, G., and Herrera-Viedma, E. (2021). A Linguistic Multi-Criteria Decision Making Methodology for the Evaluation of Tourist Services Considering Customer Opinion Value. Applied Soft Computing, 101, Article 107045. https://doi.org/10.1016/j.asoc.2020.107045

Chen, Y., Liu, X., and Wang, Q. (2024). Real-Time Analytics in Retail: Applications and Challenges. International Journal of Production Economics, 250, Article 123456.

Chen, Y., Liu, Y., and Zhou, W. (2023). Predictive Analytics in Business: Recent Advances and Future Directions. Journal of Predictive Analytics, 8(2), 167–185.

Fu, T., Ouyang, Z., Yang, Z., and Liu, S. (2020). A Product Ranking Method Combining the Features–Opinion Pairs Mining and Interval-Valued Pythagorean Fuzzy Sets. Applied Soft Computing, 97, Article 106803. https://doi.org/10.1016/j.asoc.2020.106803

Gao, S., Li, L., Li, W., Janowicz, K., and Zhang, Y. (2017). Constructing Gazetteers from Volunteered Big Geo-Data Based on Hadoop. Computers, Environment and Urban Systems, 61, 172–186. https://doi.org/10.1016/j.compenvurbsys.2014.02.004

Lee, S., Kang, S., Kim, J., and Yu, E. J. (2019). Scalable Distributed Data Cube Computation for Large-Scale Multidimensional Data Analysis on a Spark Cluster. Cluster Computing, 22(1), 2063–2087. https://doi.org/10.1007/s10586-018-1811-1

Li, M., Wu, J., and Zhang, Q. (2023). Synergies Between Business Intelligence, Artificial Intelligence, and Big Data Analytics: A Review. Journal of Business Analytics, 6(1), 78–92.

Lu, J., Liu, A., Song, Y., and Zhang, G. (2020). Data-Driven Decision Support Under Concept Drift in Streamed Big Data. Complex and Intelligent Systems, 6(1), 157–163. https://doi.org/10.1007/s40747-019-00124-4

Nie, Y., Cheng, Y., and Wu, J. (2017). Liquid-Propellant Rocket Engine Online Health Condition Monitoring Base on Multi-Algorithm Parallel Integrated Decision-Making. Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering, 231(9), 1621–1633. https://doi.org/10.1177/0954410016656878

Ninawe, M., Sarangpure, P., Sahare, P., Shende, D., and Bokade, K. (2025). A Short Review on Web Based Digital Notice Board. International Journal of Electrical Engineering and Computer Science, 14(1), 85–90.

Reveilhac, M., and Morselli, D. (2020). Digital Shift in Swiss Media Consumption Practices. Swiss Journal of Sociology, 46(3), 535–558. https://doi.org/10.2478/sjs-2020-0026

Vaidya, C., Chopade, S., Khobragade, P., Bhure, K., and Parate, A. (2025). Development of a Face Recognition-Based Attendance System. In Proceedings of the 12th International Conference on Emerging Trends in Engineering and Technology – Signal and Information Processing (ICETET–SIP) (pp. 1–6). IEEE. https://doi.org/10.1109/ICETETSIP64213.2025.11156883

Wang, L., Zhang, S., and Li, Y. (2024). Predictive Analytics for Financial Forecasting: A Comparative Study of Machine Learning Techniques. Expert Systems with Applications, 184, Article 115333. https://doi.org/10.1016/j.eswa.2024.115333

Zhang, H., Chen, L., and Wang, Y. (2024). Leveraging the Convergence of BI, AI, and Big Data Analytics for Enhanced Decision-Making: A Case Study Approach. Decision Support Systems, 150, Article 113460. https://doi.org/10.1016/j.dss.2024.113460