Added LC-Quad dataset

LC-QuAD/combiner.py

+import sys
+import json
+from queue import Queue
+from threading import Thread
+from tqdm import tqdm
+import time
+
+
+combined = {}
+
+with open("test-data.json", 'r', encoding='utf-8') as json_file:
+    data = json.load(json_file)
+
+    pbar = tqdm(total=len(data['questions']))
+
+    for i in range(len(data['questions'])):
+        question = data['questions'][i]
+        combined[question["corrected_question"]] = question["sparql_query"]
+        pbar.update(1)
+    pbar.close()
+
+with open("train-data.json", 'r', encoding='utf-8') as json_file:
+    data = json.load(json_file)
+
+    pbar = tqdm(total=len(data['questions']))
+
+    for i in range(len(data['questions'])):
+        question = data['questions'][i]
+        combined[question["corrected_question"]] = question["sparql_query"]
+        pbar.update(1)
+    pbar.close()
+
+
+print(len(combined))
+
+out = {
+    "questions": []
+}
+i = 1
+for question, query in combined.items():
+    out["questions"].append({
+        "id": str(i),
+        "question": [{
+            "language": "en",
+            "string": question
+        }],
+        "query":{
+            "sparql": query
+        }
+    })
+    i += 1
+
+with open("combined.json", 'w', encoding='utf-8') as output:
+    json.dump(out, output)