Update README.md
Browse files
README.md
CHANGED
|
@@ -1,11 +1,12 @@
|
|
| 1 |
---
|
| 2 |
license: mit
|
|
|
|
| 3 |
task_categories:
|
| 4 |
- text-classification
|
| 5 |
- text-generation
|
| 6 |
language:
|
| 7 |
- en
|
| 8 |
-
pretty_name: Relation Extraction for Inverse Relations
|
| 9 |
size_categories:
|
| 10 |
- 1K<n<10K
|
| 11 |
tags:
|
|
@@ -15,6 +16,21 @@ tags:
|
|
| 15 |
- inverse-relations
|
| 16 |
---
|
| 17 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
# Relation Extraction for Inverse Relations
|
| 19 |
|
| 20 |
This dataset is designed to evaluate the robustness of Relation Extraction (RE) models, with a specific focus on **Inverse Relations** and entity substitution. It is derived from the **FewRel** and **TekGen** benchmarks (from Text2Bench).
|
|
@@ -23,8 +39,9 @@ This dataset is designed to evaluate the robustness of Relation Extraction (RE)
|
|
| 23 |
|
| 24 |
The primary goal of this dataset is to test whether models can recognize a relationship regardless of the direction or the surface form of the entities. It includes:
|
| 25 |
1. **Inverse Mapping**: Every sample includes both the forward relation (e.g., *Mother*) and its logical inverse (e.g., *Child*).
|
| 26 |
-
|
| 27 |
-
|
|
|
|
| 28 |
### Relation Examples
|
| 29 |
| Forward Relation | Inverse Relation |
|
| 30 |
| :--- | :--- |
|
|
@@ -37,18 +54,17 @@ The primary goal of this dataset is to test whether models can recognize a relat
|
|
| 37 |
|
| 38 |
The dataset is provided in JSON format and split into four main distributions:
|
| 39 |
|
| 40 |
-
* **Original FewRel
|
| 41 |
-
* **
|
| 42 |
-
* **Relation Metadata**: Supplemental files providing human-readable definitions for Wikidata P-IDs
|
| 43 |
|
| 44 |
### Key Fields
|
| 45 |
|
| 46 |
* `tokens`: The input text (tokenized or raw string) in FewRel.
|
| 47 |
-
* `sent`: The input sentence (raw string) in Tekgen
|
| 48 |
* `head` / `tail`: The subject and object entities.
|
| 49 |
* `head_to_tail`: The ID/Name of the forward relation.
|
| 50 |
* `tail_to_head`: The ID/Name of the inverse relation.
|
| 51 |
-
* `artificial_data`: A mapping showing which `Original` entities were changed to which `
|
| 52 |
|
| 53 |
## Usage
|
| 54 |
|
|
@@ -59,7 +75,7 @@ You can load this dataset for tasks such as:
|
|
| 59 |
### Repository
|
| 60 |
|
| 61 |
For scripts and more technical details on how the artificial data was generated, visit the official repository:
|
| 62 |
-
|
| 63 |
|
| 64 |
## Citation
|
| 65 |
|
|
|
|
| 1 |
---
|
| 2 |
license: mit
|
| 3 |
+
|
| 4 |
task_categories:
|
| 5 |
- text-classification
|
| 6 |
- text-generation
|
| 7 |
language:
|
| 8 |
- en
|
| 9 |
+
pretty_name: Relation Extraction for Inverse Relations FewRel
|
| 10 |
size_categories:
|
| 11 |
- 1K<n<10K
|
| 12 |
tags:
|
|
|
|
| 16 |
- inverse-relations
|
| 17 |
---
|
| 18 |
|
| 19 |
+
<script type="application/ld+json">
|
| 20 |
+
{
|
| 21 |
+
"@context": {
|
| 22 |
+
"@language": "en",
|
| 23 |
+
"@vocab": "https://schema.org/",
|
| 24 |
+
"cr": "http://mlcommons.org/croissant/",
|
| 25 |
+
"sc": "https://schema.org/"
|
| 26 |
+
},
|
| 27 |
+
"@type": "sc:Dataset",
|
| 28 |
+
"name": "InverseRelationsBenchmark",
|
| 29 |
+
"description": "Benchmark based on FewRel with original entity mentions and synthetic substitutions for robustness testing.",
|
| 30 |
+
"conformsTo": "http://mlcommons.org/croissant/1.0"
|
| 31 |
+
}
|
| 32 |
+
</script>
|
| 33 |
+
|
| 34 |
# Relation Extraction for Inverse Relations
|
| 35 |
|
| 36 |
This dataset is designed to evaluate the robustness of Relation Extraction (RE) models, with a specific focus on **Inverse Relations** and entity substitution. It is derived from the **FewRel** and **TekGen** benchmarks (from Text2Bench).
|
|
|
|
| 39 |
|
| 40 |
The primary goal of this dataset is to test whether models can recognize a relationship regardless of the direction or the surface form of the entities. It includes:
|
| 41 |
1. **Inverse Mapping**: Every sample includes both the forward relation (e.g., *Mother*) and its logical inverse (e.g., *Child*).
|
| 42 |
+
-- *Original data* : original_fewrel_inverse.json
|
| 43 |
+
3. **Synthetics Substitutions**: A version of the dataset where real-world entities are replaced with synthetic names (e.g., replacing "Denmark" with "Emilyfort") to test if models rely on memorized entity knowledge rather than linguistic context.
|
| 44 |
+
-- *synthetic_fewrel_inverse.json*
|
| 45 |
### Relation Examples
|
| 46 |
| Forward Relation | Inverse Relation |
|
| 47 |
| :--- | :--- |
|
|
|
|
| 54 |
|
| 55 |
The dataset is provided in JSON format and split into four main distributions:
|
| 56 |
|
| 57 |
+
* **Original FewRel**: The ground-truth sentences and relations. (*original_fewrel_inverse.json*)
|
| 58 |
+
* **Synthetic FewRel**: Sentences where entities have been substituted with synthetic placeholders. (*synthetic_fewrel_inverse.json*)
|
| 59 |
+
* **Relation Metadata**: Supplemental files providing human-readable definitions for Wikidata P-IDs (*fewrel_inverse_relations.json*)
|
| 60 |
|
| 61 |
### Key Fields
|
| 62 |
|
| 63 |
* `tokens`: The input text (tokenized or raw string) in FewRel.
|
|
|
|
| 64 |
* `head` / `tail`: The subject and object entities.
|
| 65 |
* `head_to_tail`: The ID/Name of the forward relation.
|
| 66 |
* `tail_to_head`: The ID/Name of the inverse relation.
|
| 67 |
+
* `artificial_data`: A mapping showing which `Original` entities were changed to which `Synthetic` placeholders.
|
| 68 |
|
| 69 |
## Usage
|
| 70 |
|
|
|
|
| 75 |
### Repository
|
| 76 |
|
| 77 |
For scripts and more technical details on how the artificial data was generated, visit the official repository:
|
| 78 |
+
[sefeoglu/inverserelations](https://github.com/sefeoglu/inverserelations)
|
| 79 |
|
| 80 |
## Citation
|
| 81 |
|