Machine-learning based big textual data analysis for job candidate resources extraction and matching

互联网招聘的快速增长催生了大量各种数字格式的简历和相关文件。为了理解这些大数据，该项目将通过高度准确的模型开发信息服务，该模型结合了机器学习和自然语言处理的力量，可以发现有用的模式，根据候选人的个人资料/简历自动将候选人与工作匹配，并预测未来的资源可用性。这将通过从各种文本数据源创建资源预测模型并开发可集成到合作伙伴公司现有解决方案中的服务来完成。该方法提出了一种经济高效的解决方案，利用当前的机器学习和自然语言处理技术，从大型非结构化文本数据中提取和使用资源信息。通过从简历中提取模式，可以自动识别某些职位的候选人并向潜在行业推荐。研究成果将有助于为合作伙伴公司不断扩张的小型企业自动寻找和筛选候选人，并使该公司能够发展和支持其当地就业市场，并在 3 至 5 年内造福当地以外的就业市场。

The rapid growth of Internet recruiting has fostered a huge number of resumes and related documents in various digital formats. To make sense of this big data, this project will develop information services from highly accurate models that combines power of machine learning and natural language processing, which can discover useful patterns for automatically matching candidates to jobs based on their profiles/resumes, and predicting future resource availability. This will be done by creating resource prediction models from various textual data sources and developing a service that can be integrated into the partner company's existing solutions. The proposed approach presents a cost-effective and efficient solution to extract and use resource information from large unstructured text data, by taking advantage of current ML and natural language processing technology. Candidates for certain positions can be automatically identified and recommended to potential industries by extracting patterns from resumes. The research outcomes will allow automated sourcing and screening of candidates for the partner company's expanding small business and will enable the company to grow and support its local employment markets as well as benefit the employment markets beyond local areas within 3 to 5 years.