1. Introduction

In the context of the digital wave permeating various industries, the field of human resource management has undergone significant transformations. The traditional manual recruitment model can no longer adequately meet businesses' needs for efficient and accurate talent acquisition. Against this backdrop, machine learning, with its strong data analysis and prediction capabilities, as well as its ability to efficiently sift through large candidate datasets, has introduced new methods for enterprise employee recruitment. Nowadays, big data and artificial intelligence technologies have become the key driving forces behind businesses' digital transformation efforts. Furthermore, in the wake of COVID-19, new recruitment modalities such as remote recruitment and online interviews have steadily gained popularity, further advancing the recruitment process toward data-driven and intelligent directions.

Recruitment plays a critical strategic role in the digital transformation of organizations, and compared to traditional methods, it needs to foster adaptation to the demands of digital talent and attract such talent [1]. The swift development of the Internet has gradually highlighted its advantages in recruitment, as its speed and vast amount of information break through the limitations of time and space [2]. For instance, the online recruitment platform system has features such as stronger real-time processing capabilities, lower operating costs, and easier maintenance.

In recent years, numerous scholars have engaged in multi-dimensional deliberations regarding the impact of digital technology on enterprises' employee recruitment. Wang Chun [3] viewed that constructing job demand characteristic models grounded in big data technology can not only explain the specific skill needs of different positions but also interpret the existing recruitment market. This is of substantial significance for enterprises in implementing effective talent cultivation strategies. Some scholars have also defined the specific functions of artificial intelligence in the recruitment process, including qualification assessment, shortening the candidate selection process, and even personalizing interactions with job applicants. They emphasize that AI is a key technology of the present in the employee recruitment process [4]. Ouakili [5], through the combination of quantitative analysis and interview methods, suggested that the implementation of AI-powered recruitment tools has had a positive impact on the recruitment process and overall enhanced the efficiency of recruitment. However, the current existing research has a rather scattered perspective and lacks systematic integration to combine theoretical research with enterprise practice. It is necessary to promote the development of the digital human resources field from fragmented research to a systematic one, ultimately serving the best decision-making in talent management and the scientific allocation of resources.

This research aims to comprehensively review and summarize the application methods, practical challenges, and future development prospects of machine learning in employee recruitment in the era of big data. This study reveals how machine learning can remarkably enhance recruitment efficiency and cut down on time and labor costs through functions such as automated resume screening, intelligent job-person matching, and proactive talent search. On the one hand, it provides theoretical and practical guidance for the intelligent transformation of enterprise recruitment, promoting technological innovation in the field of human resource management. On the other hand, it delves deeply into the real value of machine learning technology for enterprises in the context of employee recruitment. Addressing the pain points of traditional candidate search methods, such as low efficiency and high subjectivity.

2. Literature review

In the realm of academic exploration regarding recruitment processes, the machine learning algorithms employed for candidate search and matching draw upon fundamental theories including supervised learning, unsupervised learning, and reinforcement learning. These are integrated with natural language processing and graph neural network technologies to address crucial issues within recruitment scenarios systematically. For example, the Bayesian algorithm, through its probabilistic computations, can precisely ascertain the likelihood of a candidate embodying specific job-related characteristics. At the big data level, the data mining theory serves as the foundation, enabling the extraction of valuable information from a large amount of data. Additionally, the natural language processing theory allows the system to understand the candidate's resume text and the job description text. Under the joint effect of these theories, machine learning and intelligent network big data not only exert significant efficacy in the tasks of candidate search and matching, but also perfectly meet the personalized and efficient talent selection needs of enterprises.

Upon reviewing over twenty papers, it has been observed that the number of papers published within the past five years has exhibited an upward trend. This phenomenon reflects the pressing requirements of enterprises in the big data era. They strive to boost the efficiency and precision of candidate search and identification by leveraging machine learning and artificial intelligence technologies.

At the same time, this trend also showcases the evolution of research in this domain. It has shifted from mere theoretical exploration to a phase of continuous in-depth development and expansion towards practical applications. This development will continuously drive the innovation and transformation of enterprise recruitment models. As a result, it will offer robust support to enterprises engaged in the intense competition for talent.

In academic research on this topic, the majority of scholars tend to employ a mixed-method approach. The application of machine learning in the search and matching of enterprise candidates is intrinsically intricate. It involves multi-dimensional interactions among technology, human factors, and organizations. This complexity renders it arduous to conduct a comprehensive analysis using a single research method. Specifically, in terms of technology, it encompasses algorithm design and data processing. Regarding human factors, it involves decisions made by human resources departments and the experiences of job applicants. At the organizational level, it includes the enterprise recruitment process and corporate culture. Qualitative research methods, such as focus group interviews and case analysis, are capable of uncovering the authentic experiences and practical nuances of both enterprises and candidates. On the other hand, quantitative methods, like surveys and experiments, can summarize and validate macro-level patterns. The utilization of such a diverse array of research methods not only exposes the complex nature of this field and the underlying logic for achieving a comprehensive and in-depth understanding but also offers a valuable reference for the selection and refinement of research methods in future studies. Moreover, enables a more scientific and systematic exploration of the mysteries surrounding the application of machine learning in the search and identification of enterprise candidates in the big data era.

3. Advanced applications of machine learning in employee recruitment

3.1. Practical approach

In the digital age, optimizing the recruitment process and enhancing recruitment efficiency have become crucial concerns. To address these issues, Pessach et al. [6] proposed an analytical framework that integrates machine learning (ML) and mathematical modeling. Oihab et al. [7] analyzed how digital technologies contribute to improving the successive stages of the recruitment process by comparing five specialized recruitment technologies, namely a LinkedIn social network, an Udacity MOOC, a serious game called Reveal from L'Oreal, a chatbot called Ari from TextRecruit, and a massive data analysis matching system with Randstad. tech. Ionica et al. [8] offered empirical support for digital recruitment in the tourism industry. Their research was grounded in sustainable human resource management (SHRM) and the Technology Acceptance Model (TAM), thereby enriching the understanding of digital recruitment practices within a specific industry context.

For the emerging recruitment chatbots, Sami et al. [9] explored the experiences, expectations, and design requirements of recruiters regarding recruitment chatbots, finding that they can help recruiters attract potential applicants and communicate with them automatically, and collect basic applicant information. Akram et al. [10] proposed a human-centered technology acceptance model (HC-TAM) and employed an interview approach to explore the acceptance of recruitment chatbots by recruiters and job seekers. Recently, with the increasing prevalence of generative AI, Abdelhay et al. [11] conducted research and discovered that the automation-driven capabilities of generative AI can effectively reduce recruitment time, enhance objectivity, and minimize bias. This ultimately contributes to achieving more inclusive and efficient recruitment processes. Nevertheless, it should be noted that organizational size and familiarity with AI technology may lead to variations in the impact of generative AI on recruitment.

From an international perspective, scholars have explored digital recruitment in different organizational settings. Tessema et al. [12] focused on American multinational companies (AMNC), while Kapoor et al. [13] concentrated on Indian micro, small, and medium enterprises(MSME). Their research covered the digital technologies utilized in the recruitment processes, specific selection methods, and the challenges encountered, thereby providing a comprehensive view of digital recruitment across different economies and organizational scales.

All the aforementioned studies have verified the enhancing effect of digital technologies such as machine learning on recruitment efficiency. By focusing on diverse research subjects, including the tourism industry, multinational corporations, and small and medium-sized enterprises, emphasize that the application of technology needs to be combined with organizational characteristics and industry scenarios. They are providing multi-dimensional practical references for the digitalization of recruitment in enterprises of different scales.

3.2. Real-world challenges

In the era of big data, enterprises will inevitably encounter ethical challenges and privacy concerns brought about by big data during the recruitment process. Costa et al. [14] and Almeida et al. [15] both concur that algorithmic bias constitutes a significant problem. Costa et al. further discovered through research that AI faces difficulties in accurately evaluating soft skills and interpersonal communication capabilities. Skills such as teamwork, communication, and leadership, which are of utmost importance in recruitment, cannot be effectively gauged by AI. Instead, AI can merely make judgments based on superficial information and fails to delve into candidates' actual competencies. This situation may lead enterprises to overlook potential talents with outstanding soft skills, thereby exerting a negative impact on team collaboration and business growth. Irikefe [16] pointed out that in terms of data, privacy protection, accuracy, and completeness will all become challenges for enterprises in the recruitment process. Regarding data privacy, unauthorized use or leakage of personal data collected during recruitment without the candidate's consent violates their privacy rights. Concerning data accuracy and completeness, errors or omissions in the data can undermine the accuracy and reliability of the recruitment model.

From the employers' perspective, Zahra et al. [17] found seven challenges associated with the utilization of artificial intelligence as a recruitment tool. These include inefficient facilities and resources, ineffective data accumulation, resistance to change, distrust in technology, restrictive regulations, toxic work culture, and economic and social barriers.

On the whole, data quality issues and model complexity are among the risks and challenges that enterprises encounter when recruiting talent. Data bias may lead to unfairness in the model's treatment of specific groups, thereby affecting the fairness of the search process. Moreover, the complexity of machine learning models makes it difficult for enterprises to understand the reasons for candidate screening. At the same time, the instability of the model is also a risk that requires continuous optimization to adapt to market changes. In today's era of rapid digital development, if enterprises rely solely on models without combining human judgment, they may overlook the soft skills and potential of candidates, resulting in recruitment errors. Therefore, when using machine learning, enterprises need to pay attention to algorithm fairness, data security, and model transparency, and combine traditional recruitment methods to effectively address challenges, reduce risks, and optimize the candidate search and matching process.

3.3. Future prospects

Looking ahead, machine learning will be more deeply integrated with deep learning and big data analysis. Within the realm of human resource recruitment and selection, cooperation with automated systems will become an increasingly indispensable necessity [18]. As the landscape of AI technology continues to evolve and progress, the imperative of enhancing the fairness and transparency of recruitment will emerge as a crucial prerequisite [19]. Simultaneously, the significant role of human expertise in the recruitment process, particularly during interviews and decision-making stages, cannot be overlooked. This symbiotic human-machine collaboration represents the most human-centered development direction for both enterprises and job seekers.

From the perspective of job seekers, as artificial intelligence technology continues to develop, it needs to provide more personalized services to enhance the experience of job seekers, thereby helping companies attract top talent. In addition, the moderating effect of anthropomorphism will also strengthen the relationship between AI-driven recruitment and recruitment efficiency [20]. From the perspective of enterprises, the inevitable trend of digital transformation will promote the optimization of the employment structure [21]. Moreover, it will elevate the criteria that enterprises set for talents, especially those with technical proficiency.

In summary, machine learning and data networks will deeply reshape the future of enterprise recruitment, helping recruitment move towards a more scientific, efficient, and humanized direction.

4. Conclusion

This article focuses on machine learning and big data-driven models and provides a systematic review of the new recruitment model for enterprises. The discussion is conducted from three aspects: practical methods, risks and challenges, and future prospects. It was discovered that machine learning can greatly speed up the process of finding candidates by accurately pulling important details from many resumes and adjusting to what companies and market trends need, which helps move recruitment from just looking at resumes to actively predicting the best candidates. Moreover, through data algorithms that achieve multi-dimensional quantitative matching of people and positions, recruitment efficiency and quality can be effectively improved, mismatch costs reduced, and a solid foundation laid for enterprises to respond agilely to business needs. However, the application challenges cannot be ignored. Issues such as data security, high technical barriers, and resource limitations of small and medium-sized enterprises are all problems that need to be addressed urgently. Looking to the future, although challenges still exist, with the continuous improvement of technology and ecosystem, the recruitment model driven by machine learning and big data will reshape the talent competition pattern of enterprises, reconstruct human resource management and even the underlying logic of enterprise development, and lead the industry into a more precise and intelligent new era.