LECTOR achieves the highest keyphrase extraction performance among compared models at @15, Best and Mean evaluations on lecture slides¶

Claim · All claims
Evidence · 1 study · design r2 · q2
Currency · newest study 2025 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? In the technical evaluation on 2,255 lecture slides, LECTOR achieved the highest F1-scores at @15, Best and Mean (33.44%, 37.92%, 20.14%), outperforming TF-IDF, Binary score, TextRank and AttentionRank. → Erwin Daniel López Zapata 2025

Evidence¶

Erwin Daniel López Zapata 2025¶

Erwin Daniel López Zapata, Cheng Tang, Valdemar Švábenský, Fumiya Okubo, Atsushi Shimada. (2025). LECTOR: Summarizing E-book Reading Content for Personalized Student Support. Int J Artif Intell Educ 35. https://doi.org/10.1007/s40593-025-00478-6

q2 · i? · design · r2

Technical evaluation comparing LECTOR against TF-IDF, Binary score, TextRank and AttentionRank on keyphrase extraction from 2,255 lecture slides, with performance measured against syllabus ground-truth keyphrases; LECTOR "achieves the highest performance at @15, “Best” and “Mean”".

"Specifically, LECTOR achieves the highest performance at @15, “Best” and “Mean” with F1-scores of 33.44%, 37.92%, and 20.14%, respectively."

Discussion¶

Related Claims¶

• State-of-the-art embedding-based keyphrase models degrade on lecture slides while TF-IDF generalizes better to them — reports the opposite
• Students rated LECTOR-extracted keyphrases as more relevant than TF-IDF-extracted keyphrases in a human evaluation — related