Industry 4.0 and Business Informatics: How Neural Networks Make Business “Smarter”
The research work of Anastasia Lapshina, a 4th-year student of the Business Informatics program, was presented at the III International ISBM (Information Systems for Business Management) Conference in Bangkok. The baggage of scientific research, interesting to the international academic community, allowed the undergraduate student to present her work on a global research platform with the possibility of publication on the pages of the authoritative publishing house Springer.
The topic of the scientific publication by Anastasia Lapshina and the report presented at the conference by Anastasia’s scientific supervisor Pavel Valerievich Malyzhenkov was “Business Applications of Machine Learning within the “Industry 4.0” Concept”. Anastasia started doing this work in her 3rd year. By that time, she had experience working with machine learning as part of the "Intelligent Data Analysis" minor and in the format of a coursework related to one of the deep learning technologies - computer vision. Consideration of the practical application of computer vision technology using the example of one of the large cosmetics supermarkets became a logical continuation of the work done.
Using the case of the selected company, Anastasia examined the application of neural network training technology by processing a large array of data, as well as the impact of this technology on the organization’s architecture and, subsequently, indicators for several key metrics. The problem was solved of recognizing the emotions of customers who purchase any product in a supermarket, and how, depending on their emotions – joy, sadness, disappointment – the supermarket can correct its work. There have been no similar studies in this area that would include both the training process of models and the study of the impact of technology on the company's architecture - hence the relevance and novelty of the idea.
Anastasiya Lapshina
The work was quite painstaking, many details needed to be studied during the process. However, with each edit and thanks to the help and support of my scientific supervisor Pavel Valerievich Malyzhenkov, we managed to achieve quite good results in a few months of work - we received a well-trained neural network model, studied the impact of the application of such technology on the architecture of the supermarket. Since the implementation of such technology seems to be quite a long process, it will take some time before the supermarket starts using the proposed solution. However, I am sure that the demonstrated results will convince this company of its quality and rationality.
Pavel Malyzhenkov, Academic Supervisor of the Master's Program in Business Informatics at the National Research University Higher School of Economics – Nizhny Novgorod
The peculiarity of Anastasia Lapshina's work was that it offered an architectural analysis of the implementation of technology and its impact on the enterprise's activities. It is not very often that you come across studies describing the mechanism of how new technologies are integrated into the enterprise architecture. Often, all research is limited to how effectively, for example, a neural network recognizes certain emotions. But in business informatics, we need to go further, because this discipline is characterized by a comprehensive architectural vision of the enterprise. It is necessary to look at how the proposed technology will be integrated into the company's activities as a whole: how expensive it is, how effective and necessary it is at this stage or in the future, how prepared the staff is to use this technology, whether invited specialists are needed for this and, most importantly, how it will affect the company's performance. It's a habit of systems thinking.
– Anastasia, what was the most difficult part of your job?
– The most difficult thing in the work was probably to maintain patience. Not everything worked out the first time. A large number of difficulties were encountered not only during the training of deep learning models, but also during the study and search for the necessary articles and information, and the assessment of the impact of the models on the company's performance (marketing promotion strategy aimed at achieving specific business goals). This was all a long, painstaking process, during which maintaining patience and understanding the direction of the work were key aspects.
– What qualities does a bachelor’s student need to engage in scientific work?
– I would say that a great desire is necessary. Desire breeds discipline. Discipline breeds patience and persistence. I have always understood that I really want to reach the international level in publications, since this is very important for further admission to the master's program and the portfolio competition. When there is a clear goal, you can overcome any difficulties.
Pavel Malyzhenkov
In order to do scientific work, a student needs to be, firstly, proactive, and secondly, he needs to systematically interact with his scientific supervisor. Patience is also needed, because science is not done quickly, and you need to be ready for this. The peculiarity of any scientific research is that it is difficult to predict in advance the horizons and prospects that the obtained result will open up for you. You also need hard work and a certain persistence, and somewhere, perhaps, an element of luck.
– What stages of work are planned further? Will you participate in conferences with this topic?
– During my diploma work, I think I will encounter the topic of machine learning one way or another. Perhaps the topic of my diploma will be directly related to my coursework. Alternatively, it will be an application of another machine learning technology in another area. Or it is possible to increase the efficiency of neural network training and analyze some new emotions. I would definitely like to continue my scientific work and take part in other scientific conferences, but with new research on other topics.
– What skills does scientific work allow you to improve?
– Scientific work definitely improves patience. As I have already said, there are many difficulties in the process of writing an article. Many of them initially seem insurmountable, but there is a way out of any difficulties if you apply the necessary persistence, discipline and patience to find a solution.
Anastasia is currently planning to enroll in a master's program to continue her research in the field of computer science.