Matches in SemOpenAlex for { <https://semopenalex.org/work/W4316021318> ?p ?o ?g. }
Showing items 1 to 74 of
74
with 100 items per page.
- W4316021318 abstract "This paper presents a study on classifying questions into 10 categories based on cognitive science: verification, disjunction, concept, extent, example, comparison, cause, consequence, procedural, and judgmental. These types are more related and useful for a n educational environment. The source of the question dataset is real questions asked by users on Reddit and Yahoo Answer forums. As far as we know, this study is the first to compare several pre-training models for classifying questions of these types and provide a detailed analysis of the dataset. Specifically, we used BERT for feature extraction followed by classification with Random Forest, and performed fine-tuning on BERT, XLNet, and Roberta. The experimental results show that the model with the fine-tuning pre-trained language model achieved 0.8367 accuracies and 0.8116 F1-score. Moreover, evaluation on top-2 prediction produced by the best model, Roberta, shows significant improvement in accuracy to 0.9496. Detailed error and data analysis show some issues in the dataset, opening the possibility of improving dataset quality." @default.
- W4316021318 created "2023-01-14" @default.
- W4316021318 creator A5057308940 @default.
- W4316021318 creator A5061473902 @default.
- W4316021318 creator A5079062342 @default.
- W4316021318 date "2022-12-08" @default.
- W4316021318 modified "2023-10-16" @default.
- W4316021318 title "Educational Question Classification with Pre-trained Language Models" @default.
- W4316021318 cites W1832693441 @default.
- W4316021318 cites W1966354121 @default.
- W4316021318 cites W2019307700 @default.
- W4316021318 cites W2069032672 @default.
- W4316021318 cites W2070246124 @default.
- W4316021318 cites W2082666822 @default.
- W4316021318 cites W2098555131 @default.
- W4316021318 cites W2140789179 @default.
- W4316021318 cites W2166254292 @default.
- W4316021318 cites W2265846598 @default.
- W4316021318 cites W2470673105 @default.
- W4316021318 cites W2899277693 @default.
- W4316021318 cites W2962717047 @default.
- W4316021318 cites W3008993296 @default.
- W4316021318 cites W3011091101 @default.
- W4316021318 cites W3105625590 @default.
- W4316021318 cites W3177334331 @default.
- W4316021318 doi "https://doi.org/10.1109/icic56845.2022.10006957" @default.
- W4316021318 hasPublicationYear "2022" @default.
- W4316021318 type Work @default.
- W4316021318 citedByCount "0" @default.
- W4316021318 crossrefType "proceedings-article" @default.
- W4316021318 hasAuthorship W4316021318A5057308940 @default.
- W4316021318 hasAuthorship W4316021318A5061473902 @default.
- W4316021318 hasAuthorship W4316021318A5079062342 @default.
- W4316021318 hasConcept C111472728 @default.
- W4316021318 hasConcept C119857082 @default.
- W4316021318 hasConcept C137293760 @default.
- W4316021318 hasConcept C138885662 @default.
- W4316021318 hasConcept C154945302 @default.
- W4316021318 hasConcept C169258074 @default.
- W4316021318 hasConcept C204321447 @default.
- W4316021318 hasConcept C2776401178 @default.
- W4316021318 hasConcept C2779530757 @default.
- W4316021318 hasConcept C41008148 @default.
- W4316021318 hasConcept C41895202 @default.
- W4316021318 hasConcept C52622490 @default.
- W4316021318 hasConceptScore W4316021318C111472728 @default.
- W4316021318 hasConceptScore W4316021318C119857082 @default.
- W4316021318 hasConceptScore W4316021318C137293760 @default.
- W4316021318 hasConceptScore W4316021318C138885662 @default.
- W4316021318 hasConceptScore W4316021318C154945302 @default.
- W4316021318 hasConceptScore W4316021318C169258074 @default.
- W4316021318 hasConceptScore W4316021318C204321447 @default.
- W4316021318 hasConceptScore W4316021318C2776401178 @default.
- W4316021318 hasConceptScore W4316021318C2779530757 @default.
- W4316021318 hasConceptScore W4316021318C41008148 @default.
- W4316021318 hasConceptScore W4316021318C41895202 @default.
- W4316021318 hasConceptScore W4316021318C52622490 @default.
- W4316021318 hasFunder F4320322728 @default.
- W4316021318 hasLocation W43160213181 @default.
- W4316021318 hasOpenAccess W4316021318 @default.
- W4316021318 hasPrimaryLocation W43160213181 @default.
- W4316021318 hasRelatedWork W3107474891 @default.
- W4316021318 hasRelatedWork W3116896278 @default.
- W4316021318 hasRelatedWork W3204641204 @default.
- W4316021318 hasRelatedWork W4225360065 @default.
- W4316021318 hasRelatedWork W4282839226 @default.
- W4316021318 hasRelatedWork W4283016678 @default.
- W4316021318 hasRelatedWork W4308191010 @default.
- W4316021318 hasRelatedWork W4320483443 @default.
- W4316021318 hasRelatedWork W4322727400 @default.
- W4316021318 hasRelatedWork W4323021782 @default.
- W4316021318 isParatext "false" @default.
- W4316021318 isRetracted "false" @default.
- W4316021318 workType "article" @default.