From 7a4d0c5c7b3f32091333cf07ecad139d58587649 Mon Sep 17 00:00:00 2001
From: Hironsan <light.tree.1.13@gmail.com>
Date: Fri, 25 Feb 2022 11:00:32 +0900
Subject: [PATCH] Add export catalog for relation extraction

---
 backend/data_export/pipeline/catalog.py  |  6 +++
 backend/data_export/pipeline/examples.py | 52 ++++++++++++++++++++++++
 2 files changed, 58 insertions(+)

diff --git a/backend/data_export/pipeline/catalog.py b/backend/data_export/pipeline/catalog.py
index 58751145..3eb74b43 100644
--- a/backend/data_export/pipeline/catalog.py
+++ b/backend/data_export/pipeline/catalog.py
@@ -50,6 +50,11 @@ class IntentAndSlot(Format):
     extension = "jsonl"
 
 
+class JSONLRelation(Format):
+    name = "JSONL(relation)"
+    extension = "jsonl"
+
+
 class OptionDelimiter(BaseModel):
     delimiter: Literal[",", "\t", ";", "|", " "] = ","
 
@@ -79,6 +84,7 @@ Options.register(DOCUMENT_CLASSIFICATION, JSONL, OptionNone, examples.Category_J
 
 # Sequence Labeling
 Options.register(SEQUENCE_LABELING, JSONL, OptionNone, examples.Offset_JSONL)
+Options.register(SEQUENCE_LABELING, JSONLRelation, OptionNone, examples.ENTITY_AND_RELATION_JSONL)
 
 # Sequence to sequence
 Options.register(SEQ2SEQ, CSV, OptionDelimiter, examples.Text_CSV)
diff --git a/backend/data_export/pipeline/examples.py b/backend/data_export/pipeline/examples.py
index e72fc09b..f6ba84ed 100644
--- a/backend/data_export/pipeline/examples.py
+++ b/backend/data_export/pipeline/examples.py
@@ -54,6 +54,58 @@ Offset_JSONL = """
 {"text": "President Obama", "label": [ [10, 15, "PERSON"] ]}
 """
 
+ENTITY_AND_RELATION_JSONL = """
+{
+    "text": "Google was founded on September 4, 1998, by Larry Page and Sergey Brin.",
+    "entities": [
+        {
+            "id": 0,
+            "start_offset": 0,
+            "end_offset": 6,
+            "label": "ORG"
+        },
+        {
+            "id": 1,
+            "start_offset": 22,
+            "end_offset": 39,
+            "label": "DATE"
+        },
+        {
+            "id": 2,
+            "start_offset": 44,
+            "end_offset": 54,
+            "label": "PERSON"
+        },
+        {
+            "id": 3,
+            "start_offset": 59,
+            "end_offset": 70,
+            "label": "PERSON"
+        }
+    ],
+    "relations": [
+        {
+            "id": 0,
+            "from_id": 0,
+            "to_id": 1,
+            "type": "foundedAt"
+        },
+        {
+            "id": 1,
+            "from_id": 0,
+            "to_id": 2,
+            "type": "foundedBy"
+        },
+        {
+            "id": 2,
+            "from_id": 0,
+            "to_id": 3,
+            "type": "foundedBy"
+        }
+    ]
+}
+"""
+
 CategoryImageClassification = """
 [
     {