{"id":"W2018222148","doi":"10.1145/1860559.1860615","title":"Structure-aware topic clustering in social media","year":2010,"lang":"en","type":"article","venue":"","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"","keywords":"Computer science; Conversation; Meaning (existential); Social media; Variety (cybernetics); Set (abstract data type); Cluster analysis; Software; Data science; World Wide Web; Scale (ratio); Artificial intelligence; Sociology; Communication; 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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004138476,0.001098342,0.001850534,0.01104705,0.002226239,0.002925092,0.002321521,0.001713168,0.001237655],"category_scores_gemma":[0.01262185,0.0009755131,0.001557118,0.007677546,0.001093514,0.003275244,0.002313231,0.001158273,0.001229611],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001374345,"about_ca_system_score_gemma":0.001879304,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01171793,"about_ca_topic_score_gemma":0.01590642,"domain_scores_codex":[0.996173,0.001372281,0.0002504963,0.0008676457,0.0009790371,0.000357664],"domain_scores_gemma":[0.9919207,0.004786947,0.0007851435,0.0009923856,0.001151367,0.0003634892],"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":[0.0009686461,0.0004973378,0.01816619,0.0006854479,0.0005477275,0.0004080971,0.002654914,0.2414482,0.01602602,0.02221213,0.01846598,0.6779193],"study_design_scores_gemma":[0.00005006584,0.00004741278,0.003415648,0.00002898971,0.00007114949,0.0001345986,0.000434237,0.9603274,0.004023149,0.02724797,0.004171312,0.00004806197],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0680647,0.001818362,0.9226403,0.0005891085,0.0001469167,0.0002708585,0.0008813172,0.003679022,0.001909345],"genre_scores_gemma":[0.4868418,0.00103052,0.5020325,0.0001561781,0.0005605925,0.0004525355,0.00393848,0.0006435802,0.004343887],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01171793,"threshold_uncertainty_score":0.02329946,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01119701689460568,"score_gpt":0.2779358857674432,"score_spread":0.2667388688728375,"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."}}