{"response":{"docs":[{"system_create_dtsi":"2022-11-04T20:12:13Z","system_modified_dtsi":"2022-11-04T20:12:44Z","has_model_ssim":["Dataset"],"id":"8623j034w","accessControl_ssim":["a3e24de7-cc95-4011-a0d3-d361c1c5fd50"],"hasRelatedMediaFragment_ssim":["9z903144z"],"hasRelatedImage_ssim":["9z903144z"],"depositor_ssim":["mccabeen@ucmail.uc.edu"],"depositor_tesim":["mccabeen@ucmail.uc.edu"],"title_tesim":["Coherence_Evaluations - nws_environmental.csv "],"date_uploaded_dtsi":"2022-11-04T20:12:12Z","date_modified_dtsi":"2022-11-04T20:12:12Z","isPartOf_ssim":["admin_set/default"],"doi_tesim":["doi:10.7945/d7c6-rj84"],"geo_subject_tesim":["United States"],"time_period_tesim":["21st century"],"college_tesim":["Libraries"],"department_tesim":["Digital Scholarship Center (DSC)"],"creator_tesim":["McCabe, Erin E."],"publisher_tesim":["University of Cincinnati"],"subject_tesim":["Coherence","Natural Language Processing","News"],"language_tesim":["English"],"description_tesim":["CSV files containing the coherence scoring pertaining to datasets of:\r\nDocumentCount = 5,000\r\nCorpus = (one from) Federal Caselaw [cas] / Pubmed-Abstracts [pma] / Pubmed-Central [pmc] / News [nws]\r\nSearchTerm[s] = (one from) Earth / Environmental / Climate / Pollution / Random 5k documents of a specific corpus\r\n\r\nCoherence was scored across every combination of:\r\nTopicCount: 10-40\r\nHyperparameter-Alpha: [0.01, 0.31, 0.61, 0.91, symmetric, asymmetric]\r\nHyperparameter-Beta: [0.01, 0.31, 0.61, 0.91, automatic, symmetric]\r\n\r\nThe columns in this file include:\r\nValidation_Set: Which search term this scoring pertains to\r\nTopics: Number of topics in the model\r\nAlpha: Hyperparameter alpha selection from the 6 options above\r\nBeta: Hyperparameter beta selection from the 6 options above\r\nCoherence: The topic coherence score for the given model-row\r\nPerplexity: The perplexity score for the given model-row\r\n\r\n"],"license_tesim":["http://www.opendatacommons.org/licenses/by/1.0/"],"date_created_tesim":["2022"],"thumbnail_path_ss":"/assets/work-ff055336041c3f7d310ad69109eda4a887b16ec501f35afc0a547c4adb97ee72.png","suppressed_bsi":false,"actionable_workflow_roles_ssim":["admin_set/default-default-depositing"],"workflow_state_name_ssim":["deposited"],"member_ids_ssim":["9z903144z"],"member_of_collections_ssim":["Coherence Evaluations for 'Model of Models' Methods"],"member_of_collection_ids_ssim":["rr171z875"],"file_set_ids_ssim":["9z903144z"],"visibility_ssi":"open","admin_set_tesim":["Default Admin Set"],"sort_title_ssi":"COHERENCE_EVALUATIONS NWS_ENVIRONMENTALCSV ","human_readable_type_tesim":["Dataset"],"read_access_group_ssim":["public"],"edit_access_group_ssim":["admin"],"edit_access_person_ssim":["mccabeen@ucmail.uc.edu"],"nesting_collection__ancestors_ssim":["ng451k09w","ng451k09w/rr171z875"],"nesting_collection__parent_ids_ssim":["rr171z875"],"nesting_collection__pathnames_ssim":["ng451k09w/rr171z875/8623j034w"],"nesting_collection__deepest_nested_depth_isi":3,"_version_":1749763428416749568,"timestamp":"2022-11-17T17:00:41.468Z","score":0.00049999997},{"system_create_dtsi":"2022-11-03T16:58:49Z","system_modified_dtsi":"2022-11-17T16:56:18Z","has_model_ssim":["Dataset"],"id":"jq085m40n","accessControl_ssim":["274117d2-6c08-4a05-b55e-c8d89e8d6fcf"],"hasRelatedMediaFragment_ssim":["x346d567v"],"hasRelatedImage_ssim":["x346d567v"],"depositor_ssim":["mccabeen@ucmail.uc.edu"],"depositor_tesim":["mccabeen@ucmail.uc.edu"],"title_tesim":["Coherence_Evaluations - nws_pollution.csv"],"date_uploaded_dtsi":"2022-11-03T16:58:48Z","date_modified_dtsi":"2022-11-17T16:56:14Z","isPartOf_ssim":["admin_set/default"],"doi_tesim":["doi:10.7945/yevs-9z66"],"geo_subject_tesim":["United States"],"time_period_tesim":["21st century"],"college_tesim":["Libraries"],"department_tesim":["Digital Scholarship Center (DSC)"],"creator_tesim":["McCabe, Erin E."],"publisher_tesim":["University of Cincinnati"],"subject_tesim":["Coherence","News","Natural Language Processing"],"language_tesim":["English"],"description_tesim":["CSV files containing the coherence scoring pertaining to datasets of:\r\nDocumentCount = 5,000\r\nCorpus = (one from) Federal Caselaw [cas] / Pubmed-Abstracts [pma] / Pubmed-Central [pmc] / News [nws]\r\nSearchTerm[s] = (one from) Earth / Environmental / Climate / Pollution / Random 5k documents of a specific corpus\r\n\r\nCoherence was scored across every combination of:\r\nTopicCount: 10-40\r\nHyperparameter-Alpha: [0.01, 0.31, 0.61, 0.91, symmetric, asymmetric]\r\nHyperparameter-Beta: [0.01, 0.31, 0.61, 0.91, automatic, symmetric]\r\n\r\nThe columns in this file include:\r\nValidation_Set: Which search term this scoring pertains to\r\nTopics: Number of topics in the model\r\nAlpha: Hyperparameter alpha selection from the 6 options above\r\nBeta: Hyperparameter beta selection from the 6 options above\r\nCoherence: The topic coherence score for the given model-row\r\nPerplexity: The perplexity score for the given model-row"],"license_tesim":["http://www.opendatacommons.org/licenses/by/1.0/"],"date_created_tesim":["2022"],"thumbnail_path_ss":"/assets/work-ff055336041c3f7d310ad69109eda4a887b16ec501f35afc0a547c4adb97ee72.png","suppressed_bsi":false,"actionable_workflow_roles_ssim":["admin_set/default-default-depositing"],"workflow_state_name_ssim":["deposited"],"member_ids_ssim":["x346d567v"],"member_of_collections_ssim":["Coherence Evaluations for 'Model of Models' Methods"],"member_of_collection_ids_ssim":["rr171z875"],"file_set_ids_ssim":["x346d567v"],"visibility_ssi":"open","admin_set_tesim":["Default Admin Set"],"sort_title_ssi":"COHERENCE_EVALUATIONS NWS_POLLUTIONCSV","human_readable_type_tesim":["Dataset"],"read_access_group_ssim":["public"],"edit_access_group_ssim":["admin"],"edit_access_person_ssim":["mccabeen@ucmail.uc.edu"],"nesting_collection__ancestors_ssim":["ng451k09w","ng451k09w/rr171z875"],"nesting_collection__parent_ids_ssim":["rr171z875"],"nesting_collection__pathnames_ssim":["ng451k09w/rr171z875/jq085m40n"],"nesting_collection__deepest_nested_depth_isi":3,"_version_":1749763421166895104,"timestamp":"2022-11-17T17:00:34.554Z","score":0.00049999997}],"facets":[{"name":"human_readable_type_sim","items":[{"value":"Dataset","hits":2,"label":"Dataset"}],"label":"Human Readable Type Sim"},{"name":"creator_sim","items":[{"value":"McCabe, Erin E.","hits":2,"label":"McCabe, Erin E."}],"label":"Creator Sim"},{"name":"subject_sim","items":[{"value":"Coherence","hits":2,"label":"Coherence"},{"value":"Natural Language Processing","hits":2,"label":"Natural Language Processing"},{"value":"News","hits":2,"label":"News"}],"label":"Subject Sim"},{"name":"college_sim","items":[{"value":"Libraries","hits":2,"label":"Libraries"}],"label":"College Sim"},{"name":"department_sim","items":[{"value":"Digital Scholarship Center (DSC)","hits":2,"label":"Digital Scholarship Center (DSC)"}],"label":"Department Sim"},{"name":"language_sim","items":[{"value":"English","hits":2,"label":"English"}],"label":"Language Sim"},{"name":"publisher_sim","items":[{"value":"University of Cincinnati","hits":2,"label":"University of Cincinnati"}],"label":"Publisher Sim"},{"name":"date_created_sim","items":[{"value":"2022","hits":2,"label":"2022"}],"label":"Date Created Sim"},{"name":"member_of_collection_ids_ssim","items":[{"value":"rr171z875","hits":2,"label":"rr171z875"}],"label":"Member Of Collection Ids Ssim"},{"name":"generic_type_sim","items":[{"value":"Work","hits":2,"label":"Work"}],"label":"Generic Type Sim"}],"pages":{"current_page":1,"next_page":null,"prev_page":null,"total_pages":1,"limit_value":10,"offset_value":0,"total_count":2,"first_page?":true,"last_page?":true}}}