Q&A for those interested in the geology, meteorology, oceanography, and environmental sciences
WITH CleanTags AS ( SELECT Id, TRIM( REPLACE( REPLACE(tag.value, '>', ''), '<', '' ) ) AS tag -- Clean extra characters FROM Posts CROSS APPLY STRING_SPLIT( REPLACE( REPLACE(Tags, '><', ','), '<', '' ), ',' ) AS tag WHERE Id IN ( 59430106, 59481435, 59484278, 68691256, 69385064, 59622300,59741397, 59835749, 60198252, 60202610, 69499432, 69525476, 69538510, 69596586, 69841983, 69891610, 69965649, 70196914, 70333328, 70338012, 70590588, 70825390, 70990386, 71006411, 71043010, 71167600, 60866002, 60996324, 61243073, 61520938, 71209010, 71285825, 71441870, 71470160, 71496155, 71630891, 71680438, 71687929, 71767784, 71883746, 71885781, 72076723, 72086887, 72121192, 72146421, 72179718, 72263736, 72335489, 72467705, 72556795, 72558094, 72584612, 72652038, 72698170, 72713816, 72777309, 72792770, 72818097, 73024656, 62040659, 62083380, 62398875, 62803219, 63589819, 73083132, 73326892, 73407373, 73431551, 73456900 ) ), TagCooccurrence AS ( SELECT tag, COUNT(DISTINCT Id) AS cooccurrence_count FROM CleanTags WHERE tag != 'abc' GROUP BY tag ), TotalLargeLanguageModel AS ( SELECT COUNT(DISTINCT Id) AS total_large_language_model_count FROM CleanTags WHERE tag = 'federated-learning' ), TotalTags AS ( SELECT tag, COUNT(DISTINCT Id) AS total_tag_count FROM CleanTags GROUP BY tag ) SELECT tc.tag, (tc.cooccurrence_count / CAST(tll.total_large_language_model_count AS FLOAT)) AS tag_significance_threshold FROM TagCooccurrence tc JOIN TotalLargeLanguageModel tll ON 1=1 JOIN TotalTags tt ON tc.tag = tt.tag WHERE (tc.cooccurrence_count / CAST(tll.total_large_language_model_count AS FLOAT)) >= 0.001 -- Adjust threshold if needed ORDER BY tag_significance_threshold DESC;