Add MLCroissant metadata (croissant.json)

croissant.json

@@ -1,22 +1,39 @@
 {
   "@context": {
-    "@language": "en",
+    "@language": "tl",
     "@vocab": "https://schema.org/",
     "croissant": "http://mlcommons.org/croissant/",
     "sc": "https://schema.org/"
   },
   "@type": "sc:Dataset",
   "name": "halohalo",
-  "description": "Web crawled dataset containing multilingual text content. Generated by PhilCrawler.",
+  "description": "Web crawled dataset targeting Philippine languages: Tagalog. Sourced from 46 seed URLs. Includes heuristics for Tagalog-English code-switching (Taglish). Max pages limit: 10000. Contains 535 documents with 1,038,897 total words. Language distribution: Tagalog: 535. Generated by PhilCrawler.",
   "license": "https://creativecommons.org/licenses/by/4.0/",
   "url": "https://huggingface.co/datasets/internetoftim/halohalo",
+  "inLanguage": [
+    "tl"
+  ],
+  "statistics": {
+    "@type": "sc:DatasetStatistics",
+    "numRows": 535,
+    "totalWordCount": 1038897,
+    "languageDistribution": [
+      {
+        "language": "tl",
+        "languageName": "Tagalog",
+        "count": 535,
+        "percentage": 100
+      }
+    ]
+  },
   "distribution": [
     {
       "@type": "sc:FileObject",
       "@id": "raw-data",
       "name": "data.jsonl",
       "contentUrl": "https://huggingface.co/datasets/internetoftim/halohalo/resolve/main/data.jsonl",
-      "encodingFormat": "application/jsonlines"
+      "encodingFormat": "application/jsonlines",
+      "description": "The raw crawled data in JSONL format."
     }
   ],
   "recordSet": [