15 Extraordinary Machine Learning Research Articles

Description of the topic

Customers are the most valuable asset for any organization as their whole business productivity and growth depend upon the number of customers they have. That is why, customer satisfaction and customer retention are two of the most important factors organizations consider to grow their business operations and make a mark in the market (Huang, Kechadi and Buckley, 2012). Because of the significance customers bring to an organization, its churn rate is a big concern for them as it has a direct impact on their productive growth and reputation as well (Fujo, Subramanian and Khder, 2022). In this research, the problem of the organizations related to their customers can be discussed in detail and a suitable solution can be proposed for the prediction of customer churn for the organization so that they can make some beneficial changes in their services or products to retain the customers.

Research objectives

• To investigate the major factors that lead to the increased rate of customer churn in a business organization that deals with any product or service.
• To provide a suitable ML-based solution for the prediction of customer churn rates based on their basic details and shopping pattern.

Research questions

RQ1: What factors lead to the increased rate of customer churn within the company?

RQ2: How ML techniques can help organizations inaccurate predictions of the customer churn rate?

Research methodology

To fulfill the proposed objectives of the research, a quantitative research methodology can be used in which the experimental analysis can be performed to find the suitable solution for accurately predicting the rate of customer churn for the business organization using machine learning algorithms. To maintain the reliability and validity of the research, the data should only be gathered from reliable information sources such as academic papers, conference papers and journals. The plagiarized content and the use of pre-tested code should also be strictly avoided during the research.

References

• Huang, B., Kechadi, M.T. and Buckley, B., 2012. Customer churn prediction in telecommunications. Expert Systems with Applications, 39(1), pp.1414-1425.
• Fujo, S.W., Subramanian, S. and Khder, M.A., 2022. Customer churn prediction in the telecommunication industry using deep learning. Information Sciences Letters, 11(1), p.24.

Description of the topic

Since the Description of the topic of credit cards in the year 1950, it is heavily used by people to meet their financial needs and pays back the due amount within the time limit to avoid any extra fees. Banks and other financial organizations approve the loan application based on several factors ensuring that the applicant has sufficient assets to repay the credit amount in time (Kumar, Sharma and Mahdavi, 2021). But there are several cases of unethical accounting practices with the banks where the customers did not pay back their dues in time which causes huge losses to these organizations (Vilarino and Vicente, 2020). Considering this problem of the banks, a credit scoring system can be proposed in this research so that these banks and financial institutes can predict the worthiness of the applicant based on multiple factors such as salary, property, previous record etc.

Research objectives

• To determine the key factors based on which the creditworthiness of an applicant can be decided.
• To develop a machine learning-based model for predicting the crest worthiness of an applicant based on some significant factors such as previous record, employment type of the applicant, property etc.
• To evaluate the performance of the model developed during the research to ensure its accurate predictions of creditworthiness.

Research questions

RQ1: What are the key factors that influence the approval of a credit card application by an applicant and ensure its worthiness?

RQ2: How can the machine learning model be used to develop an intelligent system for predicting credit worthiness of the credit card applicant?

Research methodology

The proposed research can be conducted using the quantitative research methodology for obtaining more objective and reliable results. Under the quantitative research methodology, experimental analysis can be performed considering the data of multiple credit card applicants to build a machine learning model that can analyze the details of the applicants and then provide accurate predictions regarding their creditworthiness.

References

• Kumar, A., Sharma, S. and Mahdavi, M., 2021. Machine learning (ML) technologies for digital credit scoring in rural finance: A literature review. Risks, 9(11), p.192.
• Vilarino, R. and Vicente, R., 2020. Dissecting racial bias in a credit scoring system experimentally developed for the Brazilian population. arXiv preprint arXiv:2011.09865.

Description of the topic

Business organizations are rapidly moving towards technological advancements for obtaining operational efficiency and faster processing of their key performance indicators. From the manufacturing of products to their packaging and delivery, retail organizations are heavily relying on these technologies for their operations (Foead, Ghifari, Kusuma, Hanafiah and Gunawan, 2021). But, still, there are a few areas where some enhancements are required even in the technologically advanced solutions such as for finding the shortest path between two points (Kim, Pena, Moll, Bennett and Kavraki, 2017). The automated vehicles and robots used in warehouses lack the shortest or optimal path-finding capabilities which increases the time for the movement of products within the warehouse and also requires more computational resources to manage these movements. This problem with the advanced solutions used in the warehouses can be discussed in this research study and an effective solution can be identified and implemented to resolve the problem.

Research objectives

The key aims and objectives of this research study are to-

• Identify the methods that can be used for finding the shortest and optimal path between two points.
• Investigate the available methods and determine the most effective one among them from the existing research studies.
• Compare the working and performance of different search algorithms or pathfinding algorithms to determine the shortest and optimal path between two points.

Research questions

RQ1: What are the applications of pathfinding algorithms in inventory management procedures?

RQ2: How does the pathfinding algorithm be used in finding the shortest path between two points within a warehouse?

Research methodology

During this research study, a series of objectives can be fulfilled to identify the best solution for finding the optimal and shortest path between two points within the warehouse. This implies that the automated vehicles and robots can be integrated with this algorithm to obtain operational efficiency. The best method or algorithm for pathfinding can be identified based on the review of the existing research studies while the quantitative research methodology can be used in this research to implement the identified algorithm to evaluate its performance and its effectiveness for the proposed objectives of the research study.

References

• Kim, S.M., Pena, M.I., Moll, M., Bennett, G.N. and Kavraki, L.E., 2017. A review of parameters and heuristics for guiding metabolic pathfinding. Journal of cheminformatics, 9(1), pp.1-13.
• Foead, D., Ghifari, A., Kusuma, M.B., Hanafiah, N. and Gunawan, E., 2021. A systematic literature review of A* pathfinding. Procedia Computer Science, 179, pp.507-514.

Description of the topic

Along with the countless benefits of the technology such as ease of life, faster processing and elimination of human errors, there are several issues that it has brought with it (Puri and Modi, 2020). The increased number of cyber-attacks is one of them. Cyber attackers are now equipped with some extremely powerful tools because of the advancements which can penetrate security systems with ease (Jerlin and Marimuthu., 2018). The organizations are not able to detect these attackers when they enter the system in the form of malware which leads to the complete disruption of the business operations (Baptista, Shields and Kolokotronis, 2019). To strengthen the security measures of the business organizational structures and to help them in preventing their systems from being compromised by cyber attackers, a malware detection system is proposed in this research which can accurately detect the malware before it enters the network based on the log features it is trained of.

Research objectives

• To identify the limitations of currently used malware detection systems.
• To build an advanced malware detection system using machine learning techniques.
• To evaluate the performance of the deep learning and machine learning-based malware detection systems.

Research questions

Throughout this research, investigation can be done and experiments can be done to find the answers to the following questions-

RQ1: How impactful can the malware be, once it enters the systems without getting noticed?

RQ2: How can machine learning or deep learning models be used to develop malware detection systems?

RQ3: Which machine learning or deep learning algorithm is more reliable for the detection of malware?

Research methodology

To find the answers to the questions raised during this research, quantitative research methodology can be used in which multiple machine learning and deep learning models can be implemented as part of the experimental analysis process to determine the best and most reliable model for detecting malware. The secondary data for the analysis can be gathered from the Kaggle repository and the information that can be included in the report can be gathered from some reliable sources such as Science Direct, IEEE Xplore, Emerald and Google Scholar.

References

• Puri, R.K. and Modi, P., 2020. Malware Detection System using Machine Learning.
• Jerlin, M.A. and Marimuthu, K., 2018. A new malware detection system using machine learning techniques for API call sequences. Journal of Applied Security Research, 13(1), pp.45-62.
• Baptista, I., Shiaeles, S. and Kolokotronis, N., 2019, May. A novel malware detection system based on machine learning and binary visualization. In 2019 IEEE International Conference on Communications Workshops (ICC Workshops) (pp. 1-6). IEEE.

Description of the topic

The influence of social media on human lives has been increasing rapidly in modern times with approximately 58% of the total population being registered on these platforms (Nash, 2019). The posts, the trends, the news, and the information circulated on these platforms take only a few seconds to reach every household across the globe (Ramanathan, Subramanian and Parrott, 2017). The popularity of these platforms and the user base it has made them a perfect option for business organizations to marketize their products and services through the marketing campaigns on these platforms (Kusumawati, 2019). These marketing campaigns on social media platforms help them not only in gaining more customers but also in gaining more recognition in different markets across the globe.

Research objectives

• To determine the advantages of social media platforms for business organizations.
• To analyze the data of social media advertisements to determine their impact on the decision-making process of the customers.
• To build a machine learning model for the prediction of the response of the customers after watching social media advertisements for any product or service.

Research questions

RQ1: Do social media platforms have an impact on the decision-making process of customers in the retail industry?

RQ2: How can the retail industries be used to increase their sales and customer acquisition rate by running marketing campaigns on social media platforms?

Research methodology

To determine the impact of social media platforms on the decision-making process of customers, an experimental analysis process is performed using the secondary data set of the social media advertisements from the Kaggle data repository in which the details of the customers and their decisions regarding the purchase of the product marketized with the advertisement is included. All these experiments can be done under the quantitative research methodology to find more objective answers to the questions raised during the research.

References

• Nash, J., 2019. Exploring how social media platforms influence fashion consumer decisions in the UK retail sector. Journal of Fashion Marketing and Management: An International Journal, 23(1), pp.82-103.
• Ramanathan, U., Subramanian, N. and Parrott, G., 2017. Role of social media in retail network operations and marketing to enhance customer satisfaction. International Journal of Operations & Production Management.
• Kusumawati, A., 2019. Impact of digital marketing on the student decision-making process of higher education institutions: A case of Indonesia. Journal of E-Learning and Higher Education, 1(1), pp.1-11.

Description of the topic

Technological advancement has made a beneficial impact on healthcare KPIs and several operations are simplified and advanced with these technologies such as the management of records of patients, disease detection and even in diagnosis as well (Boumaraf, et al., 2021). Image visualization is also a part of these technologies that can offer several benefits for the healthcare sector in detecting several diseases with the help of machine learning approaches (Khuriwal and Mishra, 2018). The application of image visualizations in the healthcare sector is a major focus of this research study to find a suitable machine learning approach that can be used for utilizing the image visualization results to detect the type of breast cancer among women (Wang, Khosla, Gargeya, Irshad and Beck, 2016).

Research objectives

• To investigate and explore the effectiveness and significance of image visualization techniques in the healthcare sector.
• To highlight the effectiveness of machine learning techniques in detecting the type of breast cancer based on the image visualizations of ultrasonic images of breasts.
• To identify a suitable machine learning approach that can detect the type of cancer with the highest accuracy rate using image visualization.

Research questions

Along with these objectives, there are a few questions that can be answered with the successful completion of the research study. These questions are-

RQ1: Which machine learning model is more suitable for the prediction of the type of bras cancer-based on the ultrasonic images?

RQ2: How can image visualizations be used in clinical diagnosis and treatment in the healthcare sector?

Research methodology

For conducting this research study, the quantitative methodology can be considered following which, an experimental analysis process can be performed using the data of ultrasonic images of the breasts to determine the type of cancer- benign, malignant and normal. A literature-based analysis can also be performed during this research to conduct the gap analysis in the existing research studies and to identify the methods used by another researcher for a similar purpose.

References

• Boumaraf, S., et al., 2021. Conventional machine learning versus deep learning for magnification dependent histopathological breast cancer image classification: A comparative study with the visual explanation. Diagnostics, 11(3), p.528.
• Khuriwal, N. and Mishra, N., 2018, November. Breast cancer detection from histopathological images using deep learning. In 2018 3rd international conference and workshops on recent advances and innovations in engineering (ICRAIE) (pp. 1-4). IEEE.
• Wang, D., Khosla, A., Gargeya, R., Irshad, H. and Beck, A.H., 2016. Deep learning for identifying metastatic breast cancer. arXiv preprint arXiv:1606.05718.

Description of the topic

In the year 2019, there was huge havoc in the business sector with the emergence of the Covid pandemic which forced them to shut down their operations (Bahja, M., 2020). But, this led to the emergence of e-commerce companies as people started referring to online shopping rather than visiting the stores because of safety concerns and also because of the variety they were getting on these platforms (Chauhan and Sehgal 2017). Upon analyzing the profit companies get from their online stores, almost every organization began to start their e-commerce stores to get more customers (Firake and Patil, 2015). Customer satisfaction and customer demands became the major priorities of these e-commerce websites because of the increased competition in the market. In this research study, the demands and opinions of the customers can be analyzed based on their reviews using the sentiment analysis process.

Research objectives

• To investigate the methods that can be used for analyzing the opinions and demands of the customers for e-commerce companies.
• To analyze the sentiments of people based on their reviews of the products listed on e-commerce websites.
• To evaluate the effectiveness of natural language processing in analyzing the sentiments of the people based on their reviews.

Research questions

RQ1: How does the sentiment analysis process can be used for identifying the demands and preferences of the people for any product or service?

RQ2: How important is customer satisfaction and their opinions for the improved operations of an e-commerce company?

Research methodology

A quantitative research methodology can be used in this research to meet the defined objectives in which the experimental analysis process can be performed using the data of reviews of eBay products. The sentiment analysis process can be performed using NLP for which the data of reviews are gathered from their official website.

References

• Chauhan, C. and Sehgal, S., 2017, May. Sentiment analysis on product reviews. In 2017 International Conference on Computing, Communication and Automation (ICCCA) (pp. 26-31). IEEE.
• Firake, V.R. and Patil, Y.S., 2015. Survey on CommTrust: multi-dimensional trust using mining e-commerce feedback comments. Proc. Int. J. Innovative Res. Comput. Commun. Eng. IJIRCCE, 3(3).
• Bahja, M., 2020. Natural language processing applications in business. E-Business-Higher Education and Intelligence Applications.

Description of the topic

The increased cases of forest fires over the last few years have been a threat to the environment and the well-being of humans as well. With the number of forest fire cases being registered every year it is predicted that half of the world’s forest can be destroyed by 2030 (Dutta and Ghosh, 2021). Early detection of the signs of fire can help in preventing the fire from causing much damage and it can be controlled. But, because of the lack of such techniques for early detection of forest fires, the number of these calamities is increasing (Sherstyuk, Zharikova and Sokol, 2018). In this research, a suitable method for the early detection of fire signs can be developed using image classification techniques.

Research objectives

• To illustrate the negative impact of increasing forest fires on the environment and also on human lives as well.
• To train the ML-based model with the raw satellite images of the forest fire and their early signs so that it can detect the forest fires early and alert the concerned authorities.

Research questions

RQ1: What are the potential benefits of using image classification in the detection of forest fires?

RQ2: How image classification process can be done using deep learning models for the early detection of forest fires?

Research methodology

This research study can be conducted using the quantitative research methodology in which the deep learning techniques can be used to perform the experiments for building a forest fire detection system based on the signs detected from the images gathered from sensors. The raw satellite images of the forest fires can be used as secondary data for the analysis process and the deep-learning CNN model can be trained to fulfill the proposed objective of the research.

References

• Dutta, S. and Ghosh, S., 2021, April. Forest fire detection using the combined architecture of separable convolution and image processing. In 2021 1st International Conference on Artificial Intelligence and Data Analytics (CAIDA) (pp. 36-41). IEEE.
• Sherstjuk, V., Zharikova, M. and Sokol, I., 2018, April. Forest fire-fighting monitoring system based on UAV team and remote sensing. In 2018 IEEE 38th International Conference on Electronics and Nanotechnology (ELNANO) (pp. 663-668). IEEE.

Description of the topic

The volatility of cryptocurrency and the stock market is a concern for all the investors who have invested their money in digital assets and who are planning to invest in the future (Ramya, 2022). This concern of the investors is considered and explored in this research to provide a suitable solution for predicting the price pattern of the cryptocurrency to decrease the risk of financial losses to the investors.

Research objectives

• To identify the factors based on which the price of the cryptocurrencies can be predicted based on literature-based analysis.
• To evaluate the performance of machine learning models for predicting the future price pattern of the cryptocurrencies.

Research questions

RQ1: On which factors, can the price of cryptocurrencies be predicted for the future?

RQ2: How do machine learning techniques help in predicting the future price pattern of cryptocurrencies?

Research methodology

To meet the proposed objectives of the research, quantitative research methodology can be considered under which an experimental analysis process can be performed using the previous data of the crypto currency’s price to analyze the pattern and train the model to provide accurate predictions for the future prices (Shahbazi and Byun, 2021).

References

• Ramya, N., 2022, April. Crypto-Currency Price Prediction using Machine Learning. In 2022 6th International Conference on Trends in Electronics and Informatics (ICOEI) (pp. 1455-1458). IEEE.
• Shahbazi, Z. and Byun, Y.C., 2021. Improving the cryptocurrency price prediction performance based on reinforcement learning. IEEE Access, 9, pp.162651-162659.

Description of the topic

The unwanted and unknown email messages received from unknown or untrusted recipients are classified as spam messages (Mansoor, Jayasinghe and Muslim, 2021). Some of these messages come with attached malicious links that can disrupt the system’s processes once they get access to the network (Aski and Sourati, 2016). In this research study, the potential impact of spam messages on the security of the system can be discussed and the effective method for its detection and classification can be developed.

Research objectives

• To identify the negative impacts of spam messages or emails on the security of the systems within an organization or for any individual.
• To explore the current measures used for the detection of spam emails in business organizations.
• To evaluate the performance of Machine learning models for the detection of spam messages on the bases of the text used in these emails.

Research questions

RQ1: To what extent, spam messages can pose a threat to the security of a system?

RQ2: How machine learning algorithms are used for the classification and detection of spam messages?

Research methodology

To get more reliable and objective results during the research, the quantitative methodology can be preferred. Following this methodology, an experimental analysis process can be performed in which both the spam and ham (not spam) emails can be analyzed to build a model that can understand the pattern and text used in both types of emails.

References

• Mansoor, R.A.Z.A., Jayasinghe, N.D. and Muslam, M.M.A., 2021, January. A comprehensive review of email spam classification using machine learning algorithms. In 2021 International Conference on Information Networking (ICOIN) (pp. 327-332). IEEE.
• Aski, A.S. and Sourati, N.K., 2016. A proposed efficient algorithm to filter spam using machine learning techniques. Pacific Science Review A: Natural Science and Engineering, 18(2), pp.145-149.

Description of the topic

Drowsiness among drivers can be due to incomplete sleep, driving for long hours, fatigue etc. This can deviate the focus of the driver from driving and can result in dangerous road accidents (Salman, Rashid, Roy, Ahsan, and Siddique, 2021). The number of road accidents due to the drowsiness of drivers is in huge numbers and is a major concern for the authorities (Costa, M., Oliveira, D., Pinto, S. and Tavares, A., 2019). More details about the problem and its impact on society can be explored in the research and the solution to the problem can be identified.

Research objectives

• To identify the key factors that lead to drowsiness among drivers.
• To highlight the impact of the increased case of driver’s drowsiness on road safety.
• To find an automated approach for the detection of driver’s drowsiness.

Research questions

RQ1: What are the factors that lead to a driver’s drowsiness?

RQ2: How the AI-based approaches can be used for detecting drowsiness among the drivers?

Research methodology

Quantitative research methodology can be more suitable for conducting this research in which experimental analysis can be performed to build an AI-based model for detecting drowsiness among drivers using the data of different facial expressions of drivers in both normal and drowsy scenarios. Literature-based analysis can also be performed to determine the key factors that lead to drowsiness and to identify the gap in existing studies related to driver drowsiness detection.

References

• Salman, R.M., Rashid, M., Roy, R., Ahsan, M.M. and Siddique, Z., 2021. Driver drowsiness detection using ensemble convolutional neural networks on YawDD. arXiv preprint arXiv:2112.10298.
• Costa, M., Oliveira, D., Pinto, S. and Tavares, A., 2019. Detecting driver’s fatigue, distraction and activity using a non-intrusive ai-based monitoring system. Journal of Artificial Intelligence and Soft Computing Research, 9(4), pp.247-266.

Description of the topic

In the retail industry, the demand and sales forecasting processes have paramount importance and help them in maintaining their inventories in a better way (Kharfan, Chan and Firdolas, 2021). Appropriate forecasting of demands in retail organizations leads to improper inventory management issues and results in the demanding products of the company getting out of stock (Huber and Stuckenschmidt, 2020). Further details about the problem are explored during the research and a machine learning-based solution is proposed that can be used by the companies for accurate predictions of the demands of the products in different locations based on their previous sales.

Research objectives

The main objectives of this research are to illustrate the potential benefits of demand forecasting for retail organizations and also to provide a machine learning-based model for the prediction of demands for retail organizations based on their previous sales. Apart from these, the research also aimed at evaluating the performance of multiple machine learning models for demand forecasting based on their accuracy score.

Research questions

RQ1: How does demand forecasting help in better inventory management for a retail organization?

Research methodology

Demand forecasting has several benefits for the retail organization and to provide a suitable method for the forecasting of demands, this research can be conducted following the quantitative research methodology under which the experimental analysis can be performed using the sales data of a retail organization to build a machine learning model forecasting the future demands accurately.

References

• Kharfan, M., Chan, V.W.K. and Firdolas Efendigil, T., 2021. A data-driven forecasting approach for newly launched seasonal products by leveraging machine-learning approaches. Annals of Operations Research, 303(1-2), pp.159-174.
• Huber, J. and Stuckenschmidt, H., 2020. Daily retail demand forecasting using machine learning with an emphasis on calendric special days. International Journal of Forecasting, 36(4), pp.1420-1438.

Description of the topic

Along with the news channels, newspapers and news articles on the web, the social media websites such as Twitter have become another platform for people to share their thoughts and news about any topic (Monti, Frasca, Eynard, Mannion, and Bronstein, 2019). But, the credibility of social media news is questionable as several people try to spread fake news on these platforms resulting in a negative impact on society (Nasir, Khan. and Varlamis, 2021). In this research, the negative impact of fake news posted on social media platforms on society is discussed and a machine learning-based solution can be developed to detect these frauds and fake news.

Research objectives

• To identify the negative influence of fake and fraudulent news on society and the people.
• To build a machine learning-based model for the detection of fake news.
• To evaluate the performance of the implemented models using metrics such as root mean square error, accuracy score and mean absolute error.

Research questions

RQ1: What are the harmful influences of fake news on people and society?

RQ2: How can ML-based models be used for the classification of fraud and fake news on social media platforms?

RQ3: What potential advantages can the government authorities leverage with the fake news detection algorithm?

Research methodology

There are multiple objectives of this research study that can be fulfilled by using the quantitative research methodology under which an experimental analysis can be performed to build a machine learning model for fake news detection and to evaluate its performance. The data for the analysis can be gathered from Twitter which is used by most people for sharing any kind of news.

References

• Monti, F., Frasca, F., Eynard, D., Mannion, D. and Bronstein, M.M., 2019. Fake news detection on social media using geometric deep learning. arXiv preprint arXiv:1902.06673.
• Nasir, J.A., Khan, O.S. and Varlamis, I., 2021. Fake news detection: A hybrid CNN-RNN based deep learning approach. International Journal of Information Management Data Insights, 1(1), p.100007.
• Trueman, T.E., Kumar, A., Narayanasamy, P. and Vidya, J., 2021. Attention-based C-BiLSTM for fake news detection. Applied Soft Computing, 110, p.107600.

Description of the topic

Stock markets are highly volatile markets where the prices can increase or decrease at any time at any rate (Nabipour, M., Nayyeri, P., Jabani, H., Shahab, S. and Mosavi, A., 2020). Investors face a lot of financial losses because of the volatile nature of the market and thus require a solution of the accurate forecasting of the future pattern of the stock prices (Patil, P., Wu, C.S.M., Potika, K. and Orang, M., 2020). To resolve this problem, multiple machine learning and deep learning models can be implemented during this research to find the most reliable and accurate model for the forecasting of future stock prices based on the previous price pattern.

Research objectives

• To implement multiple machine learning and deep learning models for forecasting future stock prices.
• To compare the performance of all the models based on their accuracy score and prediction results using the test data.

Research questions

RQ1: What factors can be considered while analyzing the price pattern of the stocks to predict future prices?

RQ2: How can machine learning techniques help investors accurately predict the future price patterns of the stock?

Research methodology

A suitable method for forecasting the future price pattern of stocks can be identified by following the quantitative research methodology in which a comparative analysis of machine learning and deep learning models can be performed to find a reliable and more accurate model for the forecasting of stock prices. Secondary data for the analysis process can be gathered from the Kaggle data set repository.

References

• Nabipour, M., Nayyeri, P., Jabani, H., Shahab, S. and Mosavi, A., 2020. Predicting stock market trends using machine learning and deep learning algorithms via continuous and binary data; a comparative analysis. IEEE Access, 8, pp.150199-150212.
• Patil, P., Wu, C.S.M., Potika, K. and Orang, M., 2020, January. Stock market prediction using an ensemble of graph theory, machine learning and deep learning models. In Proceedings of the 3rd International Conference on Software Engineering and Information Management (pp. 85-92).

Description of the topic

Some of the popular social media platforms such as Facebook, Twitter and Instagram have millions of users worldwide which makes them a suitable marketing tool for business organizations to popularize their products and services (Pourkhani, Abdipour, Baher and Moslehpour, 2019). The marketing campaigns run on social media platforms can have several benefits for them in enhancing their productive growth which can be explored throughout this research study. Also, a method to identify the impact of these campaigns on their business processes can be provided.

Research objectives

• To investigate the application and benefits of social media platforms for business growth.
• To identify and implement suitable machine learning techniques to determine the impact of social media campaigns on the sales of a company.

Research questions

RQ1: Are social media platforms and social media marketing an effective way for business organizations for enhancing their productivity or their growth?

Research methodology

The proposed objectives of the research can be fulfilled using the quantitative research methodology in which statistical analysis can be performed to compare the sales records of the company before and after they invested in the social media marketing campaigns (Aral, S., Dellarocas, C. and Goddess, D., 2013). The secondary data for the analysis process can be gathered from the Kaggle data set repository while the other information included in the research study can be gathered from some reliable sources such as IEEE Xplore, Science Direct, Emerald etc.

References

• Pourkhani, A., Abdipour, K., Baher, B. and Moslehpour, M., 2019. The impact of social media in business growth and performance: A scientometrics analysis. International Journal of Data and Network Science, 3(3), pp.223-244.
• Aral, S., Dellarocas, C. and Goddess, D., 2013. Description of the topic to the special issue—social media and business transformation: a framework for research. Information systems research, 24(1), pp.3-13.