{"id":"W4408834179","doi":"10.1051/itmconf/20257605013","title":"Natural Language Processing Techniques for Information Retrieval Enhancing Search Engines with Semantic Understanding","year":2025,"lang":"en","type":"article","venue":"ITM Web of Conferences","topic":"Topic Modeling","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Computer science; Information retrieval; Search engine; Semantic search; Natural language processing; Human–computer information retrieval; Artificial intelligence","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":[0.003553513,0.0007395279,0.0008326141,0.003510793,0.0007041107,0.002559713,0.001276222,0.001150716,0.003768481],"category_scores_gemma":[0.009613292,0.0005533848,0.001767116,0.002803039,0.001403154,0.007414951,0.001886061,0.001708703,0.002039588],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001069218,"about_ca_system_score_gemma":0.001055817,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00168798,"about_ca_topic_score_gemma":0.002588207,"domain_scores_codex":[0.9979333,0.0009169885,0.0001800359,0.0002539011,0.0006313116,0.00008457031],"domain_scores_gemma":[0.9970846,0.001713035,0.0002278956,0.0005194891,0.000412398,0.00004261849],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001500322,0.0002616999,0.001146301,0.001550128,0.0001821779,0.0002400162,0.001458509,0.04438,0.02336593,0.4645993,0.01570873,0.4469572],"study_design_scores_gemma":[0.00006793014,0.0001181701,0.0006032868,0.0001486864,0.000121182,0.0004500465,0.0004426065,0.4237481,0.01202785,0.5035173,0.05867743,0.00007727894],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.004623831,0.001447203,0.9872668,0.0011284,0.00007116089,0.0001510226,0.0001705881,0.0009246555,0.004216403],"genre_scores_gemma":[0.09765393,0.001819648,0.8953046,0.0004785472,0.0002016834,0.0003024287,0.0004843576,0.0002073539,0.003547465],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.003768481,"threshold_uncertainty_score":0.01879299,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02692447611721513,"score_gpt":0.2920163499823629,"score_spread":0.2650918738651478,"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."}}