{"id":"W4406064005","doi":"10.1016/j.ipm.2024.104058","title":"A semantic framework for enhancing pseudo-relevance feedback with soft negative sampling and contrastive learning","year":2025,"lang":"en","type":"article","venue":"Information Processing & Management","topic":"Domain Adaptation and Few-Shot Learning","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"York University; Western University","funders":"Natural Science Foundation of Hubei Province; National Natural Science Foundation of China; Hubei Provincial Department of Education","keywords":"Relevance (law); Computer science; Sampling (signal processing); Natural language processing; Psychology; Artificial intelligence; Linguistics; Political science; Philosophy; Computer vision","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":[0.0004231409,0.0001687784,0.0001715745,0.0002469881,0.0005978916,0.0009027244,0.0002370893,0.00004859668,0.000002416616],"category_scores_gemma":[0.000283401,0.0001561593,0.00002599321,0.000569923,0.00005498348,0.002284939,0.000130393,0.0002126533,0.00001022391],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006947823,"about_ca_system_score_gemma":0.0000663158,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000003739245,"about_ca_topic_score_gemma":0.000004513387,"domain_scores_codex":[0.9988455,0.00002816169,0.0003606462,0.0002449017,0.0002370568,0.0002837514],"domain_scores_gemma":[0.9989125,0.0003517902,0.0003173737,0.0001366416,0.0002321397,0.00004961486],"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":[0.00006912414,0.00001564491,0.0001645084,0.0008553843,0.00006674846,9.993282e-7,0.01110982,0.02504327,0.00001854078,0.1509036,0.00002562831,0.8117267],"study_design_scores_gemma":[0.002039955,0.0001283586,0.00554037,0.002739382,0.00006542578,0.00000584621,0.01324622,0.919803,0.0002975384,0.04003355,0.01561161,0.0004887024],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.00189508,0.00007935504,0.9928587,0.0005832524,0.0001005386,0.0005318584,5.190608e-7,0.0002406189,0.003710096],"genre_scores_gemma":[0.6105886,0.00001558444,0.3880647,0.0009168838,0.00001178727,0.00009114832,0.000003904305,0.000005743999,0.0003016639],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8947598,"threshold_uncertainty_score":0.8704993,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01007160982246063,"score_gpt":0.2650852752459006,"score_spread":0.25501366542344,"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."}}