{"id":"W3087421864","doi":"10.1007/s40593-020-00211-5","title":"Automated Essay Scoring and the Deep Learning Black Box: How Are Rubric Scores Determined?","year":2020,"lang":"en","type":"article","venue":"International Journal of Artificial Intelligence in Education","topic":"Topic Modeling","field":"Computer Science","cited_by":85,"is_retracted":false,"has_abstract":false,"ca_institutions":"Athabasca University","funders":"","keywords":"Rubric; Grading (engineering); CONTEST; Computer science; Artificial intelligence; Deep learning; Machine learning; Natural language processing; Mathematics education; Psychology","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006228437,0.000104395,0.0001595542,0.0002094924,0.00006191152,0.0004896723,0.0009047674,0.00004067421,0.000007442518],"category_scores_gemma":[0.00119534,0.0000826184,0.00005996983,0.0002701936,0.00010145,0.0007671379,0.0001351798,0.0003292757,0.00000861183],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001110215,"about_ca_system_score_gemma":0.0002036223,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00003744743,"about_ca_topic_score_gemma":0.00002856862,"domain_scores_codex":[0.9985848,0.0001464803,0.0005165548,0.0001970653,0.0004204754,0.0001346324],"domain_scores_gemma":[0.9985614,0.0002770167,0.0005111588,0.0001101372,0.0004605346,0.00007969553],"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":[0.0001396763,0.0001253697,0.01169051,0.00002120904,0.00004882248,0.00004125393,0.02300573,0.06156796,0.001285456,0.04455195,0.00005536872,0.8574667],"study_design_scores_gemma":[0.00008378526,0.00003896419,0.002025316,0.000141464,0.000006303317,0.00005946081,0.002722776,0.9786013,0.002414017,0.01367855,0.0001311143,0.00009696152],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4283297,0.0002833173,0.54898,0.02077051,0.001440834,0.0000929286,1.85179e-7,0.00004044186,0.0000620887],"genre_scores_gemma":[0.9877716,0.0001306528,0.01098386,0.0004512859,0.0006388904,0.000003895062,5.180945e-7,0.000006595624,0.00001269532],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9170333,"threshold_uncertainty_score":0.4721921,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03866308912839007,"score_gpt":0.3147992035990448,"score_spread":0.2761361144706547,"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."}}