Scholarlink

Content-based recommender systems are highly useful for obtaining relevant information in today’s digital environments, where information overload is common. Incorporating contextual information into these systems can yield more effective results, depending on the processing techniques and approaches applied. In this paper, we present a content-based, context-sensitive recommendation method in which sentiment analysis techniques are applied for context processing and semantic embedding models are used for textual content representation. The method was evaluated using a dataset of users and documents (questions and answers) from the QA (question-answering) domain. Several experiments were conducted with different techniques and configurations to identify the best-performing solution, whose results outperformed those reported in other studies.

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

Enhancing Context-Aware Content-Based Recommendation with Semantic Embeddings and Sentiment Analysis

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Erick Taylor Reverón,
  • Alfredo Simón-Cuevas,
  • Raciel Yera Toledo

摘要

Content-based recommender systems are highly useful for obtaining relevant information in today’s digital environments, where information overload is common. Incorporating contextual information into these systems can yield more effective results, depending on the processing techniques and approaches applied. In this paper, we present a content-based, context-sensitive recommendation method in which sentiment analysis techniques are applied for context processing and semantic embedding models are used for textual content representation. The method was evaluated using a dataset of users and documents (questions and answers) from the QA (question-answering) domain. Several experiments were conducted with different techniques and configurations to identify the best-performing solution, whose results outperformed those reported in other studies.