{"id":"W2610350176","doi":"10.1109/tkde.2017.2699965","title":"Finding Related Forum Posts through Content Similarity over Intention-Based Segmentation","year":2017,"lang":"en","type":"article","venue":"IEEE Transactions on Knowledge and Data Engineering","topic":"Topic Modeling","field":"Computer Science","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Ottawa","funders":"European Research Council; Université Paris-Saclay; European Cooperation in Science and Technology","keywords":"Segmentation; Similarity (geometry); Computer science; Set (abstract data type); Point (geometry); Information retrieval; Contrast (vision); Content (measure theory); Artificial intelligence; Image (mathematics); Mathematics","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002674398,0.001149361,0.001675641,0.009365087,0.001441713,0.002269317,0.001406417,0.001782959,0.00192856],"category_scores_gemma":[0.01348159,0.0004145298,0.001152973,0.005269438,0.001177793,0.004312216,0.001841597,0.000996811,0.0008902628],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009561367,"about_ca_system_score_gemma":0.001244904,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002843731,"about_ca_topic_score_gemma":0.004066708,"domain_scores_codex":[0.9976224,0.0006037924,0.0001863459,0.0007271293,0.000625586,0.0002346785],"domain_scores_gemma":[0.9888686,0.006355958,0.001882424,0.0007892225,0.001536212,0.0005674574],"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":[0.002161504,0.001792351,0.08143952,0.001244415,0.0004892129,0.0008340733,0.003356663,0.06261551,0.04016027,0.02054802,0.00765284,0.7777056],"study_design_scores_gemma":[0.0001047571,0.001129771,0.0341913,0.000113001,0.0002776394,0.0008167717,0.001451148,0.8939607,0.01741505,0.04501634,0.005386613,0.0001368595],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3693159,0.001528788,0.619947,0.0005222565,0.0001218608,0.0007946725,0.0007417263,0.002011589,0.005016286],"genre_scores_gemma":[0.7488068,0.0002998133,0.2461053,0.0001043779,0.0003374805,0.00035093,0.001494951,0.0001895214,0.002310916],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.009365087,"threshold_uncertainty_score":0.01414376,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0766944298027059,"score_gpt":0.3090613877724966,"score_spread":0.2323669579697907,"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."}}