{"id":"W7116357590","doi":"10.1016/j.eswa.2025.130869","title":"In-context learning enhanced by multi-perspective sequential retrieval and predictive feedback for few-shot aspect-based sentiment analysis","year":2025,"lang":"en","type":"article","venue":"Expert Systems with Applications","topic":"Sentiment Analysis and Opinion Mining","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Université de Montréal","funders":"National Key Research and Development Program of China; Sichuan Province Science and Technology Support Program; National Natural Science Foundation of China; Department of Science and Technology of Sichuan Province; Organization Department of Sichuan Provincial Party Committee; Ministry of Science and Technology of the People's Republic of China; Chinese Academy of Sciences","keywords":"Sentiment analysis; Benchmark (surveying); Parsing; Language model; Labeled data; Training set","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":[0.0009576128,0.001750788,0.001138874,0.001548005,0.0005860832,0.0008299043,0.001448605,0.001110783,0.002587808],"category_scores_gemma":[0.003831295,0.0003359823,0.00106457,0.0009881406,0.0003848042,0.002377809,0.001264868,0.001462339,0.001567761],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000587595,"about_ca_system_score_gemma":0.0009065493,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005142357,"about_ca_topic_score_gemma":0.01078656,"domain_scores_codex":[0.9991431,0.0001950017,0.00005082694,0.0003151154,0.0002003544,0.00009554914],"domain_scores_gemma":[0.9989719,0.0004462766,0.00008248976,0.0001574969,0.0002757434,0.00006612656],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0006115251,0.0007196981,0.00567023,0.0004237796,0.0002080152,0.0004806936,0.0004518025,0.04003762,0.06473479,0.002766794,0.01529264,0.8686025],"study_design_scores_gemma":[0.000045179,0.0002509814,0.001636263,0.00002761152,0.00008113605,0.0001705337,0.0001511127,0.976143,0.01246909,0.005644242,0.003347873,0.00003288614],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1506615,0.00321792,0.829284,0.0005836356,0.0003106685,0.0003573816,0.001110229,0.009449202,0.005025459],"genre_scores_gemma":[0.7127125,0.0006693873,0.2767039,0.0006759499,0.0002739578,0.0002803545,0.003643049,0.0002808798,0.004760171],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005142357,"threshold_uncertainty_score":0.01022488,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01572240722063087,"score_gpt":0.304922359146526,"score_spread":0.2891999519258952,"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."}}