Opinion mining with cosine similarity captures differences in SRL-related terminology within and across individual learners¶

Claim · All claims
Evidence · 1 study · design r2 · q1
Currency · newest study 2024 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q1 i? Similarity measures between interviewees' word use and SRL reference lexicons showed variance across interviewees, with the Emotion and Motivation constructs slightly more pronounced than the others. → Mathias Mejeh 2024

Evidence¶

Mathias Mejeh 2024¶

Mathias Mejeh, Martin Rehm. (2024). Taking adaptive learning in educational settings to the next level: leveraging natural language processing for improved personalization. Education Tech Research Dev. https://doi.org/10.1007/s11423-024-10345-1

q1 · i? · design · r2

Preliminary showcase analysis of individual interview transcripts from six 11th-grade students, applying opinion mining against SRL reference lexicons with cosine similarity; findings visualized in Fig. 5. The authors state the measure "is able to capture differences within and across individuals."

"As can be seen, there is quite some variance between the different interviewees, indicating different levels of SRL proficiency. Moreo- ver, it seems that the SRL constructs “Emotion” and “Motivation” are slightly more pro- nounced than the other ones"

Discussion¶

Related Claims¶

• Sentiment analysis shows interviewees overwhelmingly used positive emotions when describing their SRL strategies and the ALT feedback — related
• The system's generated practice trajectories align with learner skill levels at a peak cosine similarity of 0.962 ± 0.014 at the primary difficulty level — related
• Cosine similarity has opposing effects on LLM coding accuracy: lower similarity within codebooks helps, higher similarity within excerpts helps — related
• POS tagging of interviews links the term strategy with words suggesting the digital tool promotes SRL — related