Presenting an Innovative Model Based on Multi-Mode Network Analysis for Examining Connections and Predicting Future Labor Market Developments

Recently, network analysis has become one of the most popular and practical areas in data science. This technology, by analyzing complex data and identifying communicative patterns among different elements, delves into a deeper understanding of the structure, behavior, and interactions within networks and complex systems. Despite the critical importance of network analysis in data science and social research, particularly in the field of human resources, there remains a lack of comprehensive studies on this topic. In particular, domestic market data have not been fully studied and discussed. This lack of attention misses opportunities for deeper understanding and presenting innovative solutions to future challenges. Therefore, researchers need to employ modern approaches to investigate the challenges in this field and propose new solutions. This study introduces an innovative model that utilizes multimodal network analysis to enhance the understanding of labor market communications and predict its future developments.

To design the proposed network, the required primary data were collected from international classifications and the initial network was drawn based on them. Then, to ensure that the network matches the current labor market conditions, data from one of the domestic job search websites were also extracted and the network was updated accordingly. The Jaccard index was employed to quantify the connections between network elements, while the algorithms of preferential attachment, Adamic-Adar, and common neighbors were utilized for network validation. The Dijkstra algorithm was used to calculate the shortest path in the network and the term frequency-inverse document frequency metric was used for ranking.

This research sought to present a new method for designing a multimodal labor market network and described how it was updated and diversified to match the dynamic changes of the labor market. With the final network obtained, the connections between labor market elements (jobs and skills) were examined, and a method for quantifying these connections was presented. Additionally, a method for calculating the most efficient job transitions within this network was outlined, along with an approach for ranking skills according to various job levels. Finally, the network was validated with three link prediction algorithms. The results indicate that the use of the preferential attachment algorithm will be the best option for predicting the future of this network.

The innovative model presented in this research offers a powerful tool for network design and analysis in understanding the labor market. It not only provides a comprehensive overview of the current and future state of the domestic labor market but also offers practical solutions for addressing future challenges. In particular, discovering communicative patterns and predicting emerging trends enables better and faster adaptation to labor market changes, which in turn leads to the development of sustainable job opportunities and economic growth. Through a deep analysis of existing data and predicting potential developments, this model can help labor market managers, policymakers, and social analysts design more effective strategies to enhance labor market capacities and optimize human resources. Ultimately, the results demonstrated that innovative models based on network analysis can open new horizons for predicting and managing future developments.