{"id":"W1777006545","doi":"10.1111/j.1756-8765.2010.01108.x","title":"Comparing Methods for Single Paragraph Similarity Analysis","year":2010,"lang":"en","type":"article","venue":"Topics in Cognitive Science","topic":"Topic Modeling","field":"Computer Science","cited_by":51,"is_retracted":false,"has_abstract":true,"ca_institutions":"Defence Research and Development Canada","funders":"Defence Research and Development Canada","keywords":"Paragraph; Computer science; Natural language processing; Artificial intelligence; Similarity (geometry); Word (group theory); Focus (optics); Semantic similarity; Vector space model; Simple (philosophy); Domain (mathematical analysis); Information retrieval; Linguistics; Mathematics; World Wide Web","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01351839,0.001369181,0.001484692,0.0146464,0.0008740912,0.003478475,0.0030907,0.002175783,0.004060419],"category_scores_gemma":[0.07110843,0.0006224399,0.002519995,0.008431969,0.0007290847,0.006267752,0.002959777,0.001952676,0.002833761],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001253552,"about_ca_system_score_gemma":0.001197858,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00223162,"about_ca_topic_score_gemma":0.003250439,"domain_scores_codex":[0.982064,0.00842347,0.001463772,0.003014099,0.004682335,0.0003523553],"domain_scores_gemma":[0.9249852,0.05467168,0.002117573,0.00910654,0.008285484,0.0008336186],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001532902,0.0004620749,0.0214126,0.001603497,0.002235415,0.0001418147,0.001120907,0.03660296,0.007568076,0.009278634,0.01256488,0.9054762],"study_design_scores_gemma":[0.0002868889,0.000898412,0.02317914,0.0002968027,0.0006230826,0.0007957633,0.001711518,0.8941132,0.01812328,0.04406646,0.0156358,0.0002695452],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1079825,0.005480453,0.8681693,0.000371811,0.0004684169,0.0006664995,0.00219463,0.008021568,0.006644795],"genre_scores_gemma":[0.3271016,0.001242226,0.6606227,0.0001415477,0.0002610148,0.0008465176,0.006040477,0.001604873,0.002139014],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.0146464,"threshold_uncertainty_score":0.07149297,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.149237096773703,"score_gpt":0.4262469145196294,"score_spread":0.2770098177459264,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}