Design of English Intelligent Question Answering System Based on Similarity Algorithm

A novel approach has been introduced to enhance the efficiency of English intelligent question answering systems by leveraging a distance-based similarity algorithm. This innovation aims to address the current challenges of knowledge simplification and low query efficiency observed in existing systems. The proposed system was evaluated with 421 sentences and 900 words, achieving accuracy rates of 81.06% and 90.08%, respectively. The overall similarity values consistently exceeded 0.7.The system incorporates a similarity threshold (λ, 0 ≤ λ ≤ 1), and experimentation revealed that varying λ did not affect the number of similar questions. Notably, at λ = 0.8, the system exhibited the highest number of similar sentences, comprising 70.3% of the overall proportion, demonstrating the system's accuracy with distance-based similarity algorithms. Additionally, when λ = 0.9, the system achieved the highest number of similar words, exceeding 90%. In summary, the system has demonstrated high accuracy in providing concise and accurate answers to English questions posed by learners.