Upload croissant-rai.json

croissant-rai.json

@@ -0,0 +1,322 @@
+{
+  "@context": {
+    "@language": "en",
+    "@vocab": "https://schema.org/",
+    "arrayShape": "cr:arrayShape",
+    "citeAs": "cr:citeAs",
+    "column": "cr:column",
+    "containedIn": "cr:containedIn",
+    "cr": "http://mlcommons.org/croissant/",
+    "data": {
+      "@id": "cr:data",
+      "@type": "@json"
+    },
+    "dataBiases": "cr:dataBiases",
+    "dataCollection": "cr:dataCollection",
+    "dataType": {
+      "@id": "cr:dataType",
+      "@type": "@vocab"
+    },
+    "dct": "http://purl.org/dc/terms/",
+    "extract": "cr:extract",
+    "field": "cr:field",
+    "fileProperty": "cr:fileProperty",
+    "fileObject": "cr:fileObject",
+    "fileSet": "cr:fileSet",
+    "format": "cr:format",
+    "includes": "http://mlcommons.org/schema/includes",
+    "isArray": "cr:isArray",
+    "isLiveDataset": "cr:isLiveDataset",
+    "jsonPath": "cr:jsonPath",
+    "key": "cr:key",
+    "md5": "cr:md5",
+    "parentField": "cr:parentField",
+    "path": "cr:path",
+    "personalSensitiveInformation": "cr:personalSensitiveInformation",
+    "recordSet": "cr:recordSet",
+    "references": "cr:references",
+    "regex": "cr:regex",
+    "repeated": "cr:repeated",
+    "replace": "cr:replace",
+    "sc": "https://schema.org/",
+    "separator": "cr:separator",
+    "source": "cr:source",
+    "subField": "cr:subField",
+    "transform": "cr:transform",
+    "rai": "http://mlcommons.org/croissant/RAI/",
+    "prov": "http://www.w3.org/ns/prov#"
+  },
+  "@type": "sc:Dataset",
+  "distribution": [
+    {
+      "@type": "https://schema.org/FileObject",
+      "@id": "repo",
+      "name": "repo",
+      "description": "The Hugging Face git repository.",
+      "contentUrl": "https://huggingface.co/datasets/seantw/DEBATE_LLM/tree/refs%2Fconvert%2Fparquet",
+      "encodingFormat": "git+https",
+      "sha256": "https://github.com/mlcommons/croissant/issues/80"
+    },
+    {
+      "@type": "https://schema.org/FileSet",
+      "@id": "parquet-files-for-config-default",
+      "name": "parquet-files-for-config-default",
+      "containedIn": {
+        "@id": "repo"
+      },
+      "encodingFormat": "application/x-parquet",
+      "includes": "default/*/*.parquet"
+    }
+  ],
+  "recordSet": [
+    {
+      "@type": "cr:RecordSet",
+      "dataType": "cr:Split",
+      "key": {
+        "@id": "default_splits/split_name"
+      },
+      "@id": "default_splits",
+      "name": "default_splits",
+      "description": "Splits for the default config.",
+      "field": [
+        {
+          "@type": "cr:Field",
+          "@id": "default_splits/split_name",
+          "dataType": "sc:Text"
+        }
+      ],
+      "data": [
+        {
+          "default_splits/split_name": "train"
+        }
+      ]
+    },
+    {
+      "@type": "cr:RecordSet",
+      "@id": "default",
+      "description": "seantw/DEBATE_LLM - 'default' subset",
+      "field": [
+        {
+          "@type": "cr:Field",
+          "@id": "default/split",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "parquet-files-for-config-default"
+            },
+            "extract": {
+              "fileProperty": "fullpath"
+            },
+            "transform": {
+              "regex": "default/(?:partial-)?(train)/.+parquet$"
+            }
+          },
+          "references": {
+            "field": {
+              "@id": "default_splits/split_name"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "default/chat_round_order",
+          "dataType": "cr:Float64",
+          "source": {
+            "fileSet": {
+              "@id": "parquet-files-for-config-default"
+            },
+            "extract": {
+              "column": "chat_round_order"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "default/event_order",
+          "dataType": "cr:Float64",
+          "source": {
+            "fileSet": {
+              "@id": "parquet-files-for-config-default"
+            },
+            "extract": {
+              "column": "event_order"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "default/event_type",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "parquet-files-for-config-default"
+            },
+            "extract": {
+              "column": "event_type"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "default/worker_id",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "parquet-files-for-config-default"
+            },
+            "extract": {
+              "column": "worker_id"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "default/sender_id",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "parquet-files-for-config-default"
+            },
+            "extract": {
+              "column": "sender_id"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "default/recipient_id",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "parquet-files-for-config-default"
+            },
+            "extract": {
+              "column": "recipient_id"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "default/field",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "parquet-files-for-config-default"
+            },
+            "extract": {
+              "column": "field"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "default/text",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "parquet-files-for-config-default"
+            },
+            "extract": {
+              "column": "text"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "default/sliderValue",
+          "dataType": "cr:Float64",
+          "source": {
+            "fileSet": {
+              "@id": "parquet-files-for-config-default"
+            },
+            "extract": {
+              "column": "sliderValue"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "default/agreement_level",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "parquet-files-for-config-default"
+            },
+            "extract": {
+              "column": "agreement_level"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "default/validity",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "parquet-files-for-config-default"
+            },
+            "extract": {
+              "column": "validity"
+            }
+          }
+        }
+      ]
+    }
+  ],
+  "name": "DEBATE_LLM",
+  "description": "\n\t\n\t\t\n\t\tDEBATE Benchmark\n\t\n\nThis repository contains CSV files from the DEBATE project: large-scale\nhuman conversation experiments organized around controversial and\nopinion-based topics. The data consists of multi-round conversations\nbetween human participants discussing political, social, and belief-related\ntopics, following the protocol described in:\n\nChuang, Y.-S., Tu, R., Dai, C., Vasani, S., Li, Y., Yao, B., Tessler, M. H., Yang, S., Shah, D., Hawkins, R., Hu, J., & Rogers, T. T. (2026).… See the full description on the dataset page: https://huggingface.co/datasets/seantw/DEBATE_LLM.",
+  "alternateName": [
+    "seantw/DEBATE_LLM"
+  ],
+  "creator": {
+    "@type": "Person",
+    "name": "Sean Chuang",
+    "url": "https://huggingface.co/seantw"
+  },
+  "keywords": [
+    "100K - 1M",
+    "csv",
+    "Tabular",
+    "Text",
+    "Datasets",
+    "Dask",
+    "Polars",
+    "Croissant",
+    "arxiv:2510.25110",
+    "🇺🇸 Region: US"
+  ],
+  "url": "https://huggingface.co/datasets/seantw/DEBATE_LLM",
+  "conformsTo": "http://mlcommons.org/croissant/1.1",
+  "rai:dataUseCases": [
+    "Benchmarking opinion dynamics in multi-round debates and conversations.",
+    "Evaluating role-playing LLM agents against human debate trajectories and interaction patterns.",
+    "Research on argumentation, consensus building, and belief updating on controversial topics."
+  ],
+  "rai:dataLimitations": [
+    "The dataset focuses on controversial opinion-based topics and should not be treated as representative of general everyday conversation.",
+    "Some conversations are incomplete because of participant dropout.",
+    "Some entries reflect automatic submissions caused by time limits.",
+    "Empty rows can arise from concatenation of consecutive messages and should not be interpreted as literal empty participant utterances.",
+    "The benchmark is appropriate for evaluating opinion-dynamics behavior in this task setting, but not for making direct claims about real-world persuasion effectiveness or population-level political behavior."
+  ],
+  "rai:dataBiases": [
+    "Topic-selection bias: political, social, and belief-related controversies are overrepresented relative to neutral conversation.",
+    "Human-sample bias: the participant pool may not be representative of the broader population in demographics, political orientation, topic familiarity, communication style, or willingness to revise opinions.",
+    "Selection and self-selection bias: people who choose to participate in online debate studies may differ systematically from non-participants in engagement level, expressiveness, and tolerance for disagreement.",
+    "Platform and task-design bias: online experimental settings, time limits, incentives, and interface constraints may shape how participants respond and may not reflect naturalistic offline or long-term discussions.",
+    "Attrition bias: conversations with participant dropout or incomplete participation may systematically differ from fully completed conversations.",
+    "The depth-versus-breadth design intentionally allocates uneven coverage across topics."
+  ],
+  "rai:personalSensitiveInformation": [
+    "Free-text discussions and scalar opinion ratings on political, social, and belief-related topics may reveal sensitive beliefs or demographic proxies.",
+    "The public release uses worker-style identifiers; authors should explicitly state whether any direct identifiers were removed or redacted before release."
+  ],
+  "rai:dataSocialImpact": "Potential positive impact: enabling reproducible evaluation of social reasoning and opinion-dynamics behavior in LLM agents. Potential negative impact: misuse for persuasion profiling, overclaiming real-world social competence, or benchmarking politically sensitive content without adequate safeguards.",
+  "rai:hasSyntheticData": false,
+  "prov:wasDerivedFrom": [],
+  "prov:wasGeneratedBy": [
+    "Data collection: original human multi-round conversation experiments on controversial opinion-based topics; see Appendix X for collection protocol, timeframe, and participant recruitment details.",
+    "Preprocessing: raw experiment exports were cleaned and organized into raw and golden subsets; golden subsets require complete end-to-end participation, non-empty required fields, and validity-rate thresholds; see Appendix X.",
+    "Annotation / derived fields: validity labels, agreement levels, slider values, and golden-data criteria are documented in Appendix X together with quality-control procedures."
+  ]
+}