Scholarlink

Abstract <p>The paper analyzes requirements of companies in order to identify new knowledge therefrom to supplement and adjust academic programs. This task is solved by computer-aided processing of natural language texts of the requirements using machine learning methods. The Doc2Vec model was used to implement the metric of comparing the text of a discipline syllabus and company requirements. It takes into account the context of words and their meanings. Knowledge is presented in the form of a semantic network, an digraph, which is constructed from a text parse tree under semantic-syntactic rules. New knowledge is extracted for a specific discipline syllabus by comparing knowledge graphs constructed from the texts of the requirements and the syllabus. Experiments have confirmed the possibility of automatically supplementing discipline syllabi with new notions extracted from the texts of company requirements for university alumni. The inclusion of a new agent that has this role in the multi-agent training modeling system will enable the department to respond quickly to current labor market demands.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Knowledge Extraction from Texts of Company Requirements to Update University Department Discipline Syllabi

  • 期刊论文

  • 发表日期：

    2025年12月1日

查看全文
  • N. V. Meleshchenko,
  • O. I. Fedyaev

摘要

Abstract

The paper analyzes requirements of companies in order to identify new knowledge therefrom to supplement and adjust academic programs. This task is solved by computer-aided processing of natural language texts of the requirements using machine learning methods. The Doc2Vec model was used to implement the metric of comparing the text of a discipline syllabus and company requirements. It takes into account the context of words and their meanings. Knowledge is presented in the form of a semantic network, an digraph, which is constructed from a text parse tree under semantic-syntactic rules. New knowledge is extracted for a specific discipline syllabus by comparing knowledge graphs constructed from the texts of the requirements and the syllabus. Experiments have confirmed the possibility of automatically supplementing discipline syllabi with new notions extracted from the texts of company requirements for university alumni. The inclusion of a new agent that has this role in the multi-agent training modeling system will enable the department to respond quickly to current labor market demands.