Opinion Mining for Educational Video Lectures

Area of Science:

• Educational Technology
• Natural Language Processing
• Information Retrieval

Background:

• Searching for relevant educational videos is time-consuming.
• The growing demand for educational content exacerbates this challenge.
• Users often rely on limited information from hosting pages for selection.

Purpose of the Study:

• To classify user comments on educational videos as positive or negative.
• To provide users with a qualitative overview of video feedback.
• To enhance the selection process for educational video content.

Main Methods:

• Automatic identification of significant keywords from video lecture transcripts.
• Filtering user comments based on their semantic relevance to identified keywords.
• Sentiment analysis of filtered comments to determine positive or negative sentiment.

Main Results:

• A system for filtering and classifying educational video comments was developed.
• The method successfully links comments to video content through keyword analysis.
• This approach offers a more relevant and qualitative assessment of user feedback.

Conclusions:

• Automatic keyword identification and comment filtering improve the relevance of user feedback for educational videos.
• This technique aids users in making more informed decisions when selecting educational video content.
• The study enhances the discoverability and evaluation of educational video resources.