Update README.md
Browse files
README.md
CHANGED
|
@@ -1,8 +1,13 @@
|
|
| 1 |
## Dataset Card for `Text2sparql-Raw`
|
| 2 |
|
| 3 |
-
### Dataset Summary
|
| 4 |
|
| 5 |
-
`Text2Sparql-Raw` is a multilingual dataset designed for the task of translating natural language questions into SPARQL queries over the DBpedia knowledge graph. This dataset aggregates and harmonizes four widely used benchmarks in the text-to-SPARQL domain:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
|
| 7 |
It contains questions in both **English** and **Spanish**, making it suitable for multilingual and cross-lingual question answering tasks over knowledge graphs.
|
| 8 |
|
|
@@ -10,20 +15,20 @@ This is the **first version** of the dataset.
|
|
| 10 |
|
| 11 |
---
|
| 12 |
|
| 13 |
-
### Supported Tasks and Leaderboards
|
| 14 |
|
| 15 |
* **Text-to-SPARQL generation**: Given a natural language question, the task is to generate a syntactically correct and semantically accurate SPARQL query that retrieves the answer from the DBpedia knowledge graph.
|
| 16 |
|
| 17 |
---
|
| 18 |
|
| 19 |
-
### Languages
|
| 20 |
|
| 21 |
-
* English (`en`)
|
| 22 |
-
* Spanish (`es`)
|
| 23 |
|
| 24 |
---
|
| 25 |
|
| 26 |
-
### Dataset Structure
|
| 27 |
|
| 28 |
Each instance in the dataset includes:
|
| 29 |
|
|
@@ -33,7 +38,7 @@ Each instance in the dataset includes:
|
|
| 33 |
|
| 34 |
---
|
| 35 |
|
| 36 |
-
### Dataset Creation
|
| 37 |
|
| 38 |
The dataset is a **curated compilation** of four public datasets:
|
| 39 |
|
|
@@ -46,7 +51,7 @@ The merged dataset aims to provide a unified resource for training multilingual
|
|
| 46 |
|
| 47 |
---
|
| 48 |
|
| 49 |
-
### Use Cases
|
| 50 |
|
| 51 |
This dataset is intended for training and evaluating models that:
|
| 52 |
|
|
@@ -56,7 +61,7 @@ This dataset is intended for training and evaluating models that:
|
|
| 56 |
|
| 57 |
---
|
| 58 |
|
| 59 |
-
### Citation
|
| 60 |
|
| 61 |
If you use this dataset in your research, please cite:
|
| 62 |
|
|
@@ -71,7 +76,7 @@ If you use this dataset in your research, please cite:
|
|
| 71 |
|
| 72 |
---
|
| 73 |
|
| 74 |
-
### Licensing
|
| 75 |
|
| 76 |
license: apache-2.0
|
| 77 |
|
|
|
|
| 1 |
## Dataset Card for `Text2sparql-Raw`
|
| 2 |
|
| 3 |
+
### 🧾 Dataset Summary
|
| 4 |
|
| 5 |
+
`Text2Sparql-Raw` is a multilingual dataset designed for the task of translating natural language questions into SPARQL queries over the DBpedia knowledge graph. This dataset aggregates and harmonizes four widely used benchmarks in the text-to-SPARQL domain:
|
| 6 |
+
|
| 7 |
+
* **QALD (versions 1–9)**
|
| 8 |
+
* **LC-QuAD 1.0**
|
| 9 |
+
* **Orange/paraqa-sparqltotext**
|
| 10 |
+
* **julioc-p/Question-Sparql**
|
| 11 |
|
| 12 |
It contains questions in both **English** and **Spanish**, making it suitable for multilingual and cross-lingual question answering tasks over knowledge graphs.
|
| 13 |
|
|
|
|
| 15 |
|
| 16 |
---
|
| 17 |
|
| 18 |
+
### ✅ Supported Tasks and Leaderboards
|
| 19 |
|
| 20 |
* **Text-to-SPARQL generation**: Given a natural language question, the task is to generate a syntactically correct and semantically accurate SPARQL query that retrieves the answer from the DBpedia knowledge graph.
|
| 21 |
|
| 22 |
---
|
| 23 |
|
| 24 |
+
### 🌍 Languages
|
| 25 |
|
| 26 |
+
* **English** (`en`)
|
| 27 |
+
* **Spanish** (`es`)
|
| 28 |
|
| 29 |
---
|
| 30 |
|
| 31 |
+
### 📁 Dataset Structure
|
| 32 |
|
| 33 |
Each instance in the dataset includes:
|
| 34 |
|
|
|
|
| 38 |
|
| 39 |
---
|
| 40 |
|
| 41 |
+
### 🔧 Dataset Creation
|
| 42 |
|
| 43 |
The dataset is a **curated compilation** of four public datasets:
|
| 44 |
|
|
|
|
| 51 |
|
| 52 |
---
|
| 53 |
|
| 54 |
+
### 💡 Use Cases
|
| 55 |
|
| 56 |
This dataset is intended for training and evaluating models that:
|
| 57 |
|
|
|
|
| 61 |
|
| 62 |
---
|
| 63 |
|
| 64 |
+
### 📚 Citation
|
| 65 |
|
| 66 |
If you use this dataset in your research, please cite:
|
| 67 |
|
|
|
|
| 76 |
|
| 77 |
---
|
| 78 |
|
| 79 |
+
### 🪪 Licensing
|
| 80 |
|
| 81 |
license: apache-2.0
|
| 82 |
|