Construction of English and American Literature Corpus Based on Machine Learning Algorithm

Area of Science:

• Computational Linguistics
• Educational Technology
• Digital Humanities

Background:

• Corpus application in English and American literature teaching in China is nascent.
• Existing methods lack attention from educators, hindering effective literature pedagogy.
• A systematic approach to corpus construction is needed to advance literature education.

Purpose of the Study:

• To develop an automated methodology for constructing English and American literature corpora.
• To integrate keyword extraction and advanced text classification for corpus building.
• To evaluate the efficacy of the proposed automated corpus construction method.

Main Methods:

• Keyword and key phrase extraction for content identification.
• TextRank algorithm for calculating atomic event similarity and sentence selection.
• A hybrid machine learning classifier combining Support Vector Machine (SVM) and Naive Bayes (NB) for text classification.

Main Results:

• The combined SVM and NB classifier demonstrated superior performance in accuracy and recall compared to individual methods.
• Achieved optimal classification results with an accuracy of 0.87, recall of 0.9, and F-value of 0.89.
• The proposed method effectively and efficiently generates high-quality bilingual mixed web pages for corpus use.

Conclusions:

• Automated corpus construction significantly enhances English and American literature teaching resources.
• The hybrid machine learning approach offers a robust solution for accurate text classification in corpus development.
• This research provides a scalable and precise method for creating valuable digital literary resources.