Optimization of the work distribution process for managing activities in IT companies using machine learning

The purpose of this work is to optimize the work distribution process when working in an IT company using machine learning algorithms, which allows to reduce the load on the key team member - the manager, who is the master of the scrum. We consider a team in scrum according to the Takman model, highlighting 5 stages of development: formation, conflict, normalizing, executive, separation. To increase the time for micro-management within the team, it is proposed to automate routine processes using an optimization approach. It is shown that the most time-consuming part of this definition is the definition of the type of tasks. To do this, it is necessary to define the terms of reference, assign its components (tasks) to a specific type and entrust the implementation to the developer who will achieve the required result in the shortest possible time. The structure of the optimization model for the distribution of tasks between developers is considered. Preliminary classification of tasks by category, taking into account the capabilities of the team. The choice of a convolutional neural network and the use of deep machine learning for solving the classification problem has been substantiated. As the initial data when training the network at the initial stages of team development, it is proposed to use the texts of test tasks and their distribution by categories of project tasks.

1. 5 models of effective team interaction. Available at: https://habr.com/ru/company/hygger/blog/418001/ (accessed 12.01.2021).

2. Flexible Agile Process Methodology. Available at: https://intuit.ru/studies/courses/3590/832/info (accessed 20.01.2021).

3. Tuckman, B.: Developmental Sequence in Small Groups. Psychological Bulletin, 1965:384-399.

4. Tuckman, B., Jensen, M.: Stages of Small Group Development. Group and Organizational Studies, 1977:419-427.

5. Y. LeCun, B. Boser, J. S. Denker, D. Henderson, R. E. Howard, W. Hubbard and L. D. Jackel. Backpropagation Applied to Handwritten Zip Code Recognition, Neural Computation, 1(4)1989:541-551.

6. Scrum Mastery: 5 Steps to Improve Team Process. Available at: https://www.agilesocks.com/scrum-mastery-5-steps-improve-team-process/ (accessed 20.01.2021).

7. Secure the software development lifecycle with machine learning. Available at: https://www.microsoft.com/security/blog/2020/04/16/secure-software-development-lifecycle-machine-learning/ (accessed 10.01.2021).

8. Lvovich Y.E. Multiple Optimization: Theory and Applications. Voronezh: Publishing house «Kvarta». 2006.

9. Lvovich Y.E. Decision making in expert-virtual environment. Y.E. Lvovich, I.Y. Lvovich, Voronezh: PPC «Scientific book», 2010.

10. Hastie T., Tibshirani R., Friedman J. The Elements of Statistical Learning. Springer, 2001.

11. Zhang, X. Character-level convolutional networks for text classification. Xiang Zhang, Junbo Zhao, Yann LeCun. In Advances in Neural Information Processing Systems. 2015:649-657.

12. Kim, Y. Convolutional neural networks for sentence classification. Yoon Kim. IEMNLP. – 2014:1746-1751.