Add table CSV JSON conversion benchmark data

.gitattributes

@@ -58,3 +58,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 # Video files - compressed
 *.mp4 filter=lfs diff=lfs merge=lfs -text
 *.webm filter=lfs diff=lfs merge=lfs -text
+table_csv_json_conversion_gittables_benchmark/metadata.json filter=lfs diff=lfs merge=lfs -text
+table_csv_json_conversion_wdc_webtables_sample/metadata.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,10 @@
+# data
+
+Table/CSV/JSON conversion benchmark sources normalized as metadata JSON.
+
+## Folders
+
+- `table_csv_json_conversion_gittables_benchmark`: 표/CSV/JSON 변환, 1101 records
+- `table_csv_json_conversion_wdc_webtables_sample`: 표/CSV/JSON 변환, 9275 records
+
+Raw ZIP/GZ/CSV files are excluded from the repository.

SOURCE_MAPPING.md

@@ -0,0 +1,12 @@
+# Source Mapping
+
+| folder | subdomain | source | records | license note |
+| --- | --- | --- | ---: | --- |
+| `table_csv_json_conversion_gittables_benchmark` | 표/CSV/JSON 변환 | GitTables benchmark | 1101 | CC BY 4.0; table-level original licenses unavailable in benchmark subset |
+| `table_csv_json_conversion_wdc_webtables_sample` | 표/CSV/JSON 변환 | WDC WebTables 2015 sample | 9275 | public download; data license unspecified |
+
+## Full Upstream Corpus Notes
+
+- GitTables 1M primary corpus is 16.3 GB and consists of 1M parquet tables.
+- WDC complete corpus 2015 is 165 GB; relational corpus is 69 GB; English relational corpus is 69 GB.
+- This repo stores normalized benchmark-ready subsets rather than mirroring the upstream bulk archives.

crawl_summary.json

@@ -0,0 +1,27 @@
+{
+  "domain": "데이터/분석 실무",
+  "subdomain": "표/CSV/JSON 변환",
+  "total_records": 10376,
+  "record_counts_by_folder": {
+    "table_csv_json_conversion_gittables_benchmark": 1101,
+    "table_csv_json_conversion_wdc_webtables_sample": 9275
+  },
+  "record_counts_by_source": {
+    "GitTables benchmark - column type detection": 1101,
+    "WDC Web Table Corpus 2015 random sample": 9275
+  },
+  "sources": [
+    {
+      "source": "GitTables benchmark - column type detection",
+      "url": "https://zenodo.org/records/5706316",
+      "license": "CC BY 4.0",
+      "upstream_full_corpus": "GitTables 1M, 16.3 GB"
+    },
+    {
+      "source": "WDC Web Table Corpus 2015 random sample",
+      "url": "https://data.dws.informatik.uni-mannheim.de/webtables/2015-07/sample.gz",
+      "license": "data license unspecified; public download",
+      "upstream_full_corpus": "Complete corpus 165 GB; relational corpus 69 GB; English relational corpus 69 GB"
+    }
+  ]
+}

table_csv_json_conversion_gittables_benchmark/README.md

@@ -0,0 +1,8 @@
+# table_csv_json_conversion_gittables_benchmark
+
+- domain: 데이터/분석 실무
+- subdomain: 표/CSV/JSON 변환
+- records: 1101
+- source: GitTables benchmark - column type detection
+- license: CC BY 4.0; per-table original license is not exposed in this benchmark subset
+- format: `metadata.json`

table_csv_json_conversion_gittables_benchmark/metadata.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f4430e74c142d13c925c957c5c24455e391050822c622accb9c44121627369f
+size 38353362

table_csv_json_conversion_wdc_webtables_sample/README.md

@@ -0,0 +1,8 @@
+# table_csv_json_conversion_wdc_webtables_sample
+
+- domain: 데이터/분석 실무
+- subdomain: 표/CSV/JSON 변환
+- records: 9275
+- source: WDC Web Table Corpus 2015 random sample
+- license: public download; data license unspecified; extraction code license is Apache
+- format: `metadata.json`

table_csv_json_conversion_wdc_webtables_sample/metadata.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7823a23a20c8dcfbaf1fd5f6ed181831bfaa892560cc68bcc33a95571f48d945
+size 84140719