update readme
#2
by
iremddemir
- opened
README.md
CHANGED
@@ -1,3 +1,109 @@
|
|
1 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
license: gpl-3.0
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
dataset_info:
|
3 |
+
- config_name: mlsum_tr_ext
|
4 |
+
- config_name: xtinge-sum_tr_ext
|
5 |
+
- config_name: tes
|
6 |
+
configs:
|
7 |
+
- config_name: mlsum_tr_ext
|
8 |
+
data_files:
|
9 |
+
- split: train
|
10 |
+
path: MLSUM_TR_EXT/train*
|
11 |
+
- split: test
|
12 |
+
path: MLSUM_TR_EXT/test*
|
13 |
+
- split: val
|
14 |
+
path: MLSUM_TR_EXT/val*
|
15 |
+
- config_name: xtinge-sum_tr_ext
|
16 |
+
data_files:
|
17 |
+
- split: test
|
18 |
+
path: XTINGE-SUM_TR_EXT/XTINGE-SUM_TR_EXT*
|
19 |
+
- config_name: tes
|
20 |
+
data_files:
|
21 |
+
- split: test
|
22 |
+
path: TES/tes*
|
23 |
+
task_categories:
|
24 |
+
- summarization
|
25 |
license: gpl-3.0
|
26 |
---
|
27 |
+
# XTINGE Turkish Extractive Summarization Datasets
|
28 |
+
|
29 |
+
This repository hosts three datasets created for advancing Turkish extractive text summarization research: MLSUM_TR_EXT, TES, and XTINGE-SUM_TR_EXT. These datasets are designed to support the development of models capable of generating concise and relevant extractive summaries of Turkish texts.
|
30 |
+
|
31 |
+
Below is a Python example showcasing how to download and use these datasets:
|
32 |
+
|
33 |
+
```python
|
34 |
+
from datasets import load_dataset
|
35 |
+
|
36 |
+
# Load the MLSUM_TR_EXT dataset
|
37 |
+
mlsum_tr_ext = load_dataset("xtinge/turkish-extractive-summarization-dataset", "mlsum_tr_ext")
|
38 |
+
# Load the TES dataset
|
39 |
+
tes = load_dataset("xtinge/turkish-extractive-summarization-dataset", "tes")
|
40 |
+
# Load the xtinge-sum_tr_ext dataset
|
41 |
+
xtinge_sum_tr_ext = load_dataset("xtinge/turkish-extractive-summarization-dataset", "xtinge-sum_tr_ext")
|
42 |
+
|
43 |
+
```
|
44 |
+
|
45 |
+
## Dataset Details
|
46 |
+
|
47 |
+
### Dataset Description
|
48 |
+
|
49 |
+
The datasets, having a focus on Turkish text summarization, aim to advance research in this area by providing structured, annotated resources for extractive summarization tasks. These datasets are:
|
50 |
+
|
51 |
+
1. **MLSUM_TR_EXT**:
|
52 |
+
- Originates as an extension of the Turkish subset from the [MLSUM dataset](https://huggingface.co/datasets/mlsum), focusing on extractive summarization.
|
53 |
+
- Comprises articles from internethaber.com, with summaries derived from existing headlines for creating contextually rich extractive summaries.
|
54 |
+
- Sentences within these articles were selected based on their SBERT Similarity and ROUGE Scores compared to the original summaries, ensuring relevance and conciseness.
|
55 |
+
|
56 |
+
2. **TES**:
|
57 |
+
- Represents a unique collection found on [Hugging Face](https://huggingface.co/erturkerdagi/turkishExtractiveSummarization/tree/main) tailored for Turkish extractive summarization.
|
58 |
+
- Contains a variety of news articles annotated by three distinct annotators, each providing different perspectives and lengths, thus contributing to a rich set of summarization examples.
|
59 |
+
|
60 |
+
3. **XTINGE-SUM_TR_EXT**:
|
61 |
+
- Specifically developed to supplement existing resources by providing detailed sentence importance rankings within lengthy Wikipedia documents.
|
62 |
+
- Features annotations by three different annotators who meticulously ranked all sentences by importance, contributing to a comprehensive resource for studying extractive summarization.
|
63 |
+
- The annotation process considered Inter Annotator Agreement, specifically employing Krippendorff's alpha to ensure consistency and reliability in sentence importance assessments.
|
64 |
+
|
65 |
+
- **Language(s) (NLP):** Turkish
|
66 |
+
- **License:** [gpl-3.0]
|
67 |
+
|
68 |
+
## Dataset Structure
|
69 |
+
|
70 |
+
### Generic Structure Across Datasets
|
71 |
+
|
72 |
+
All three datasets share a generic structure tailored for extractive summarization tasks, comprising the following elements:
|
73 |
+
|
74 |
+
- **Title**: The title of the document or article, serving as a concise representation of the content.
|
75 |
+
- **Sentences**: The body of the text, split into sentences. This segmentation facilitates the identification of individual sentences that contribute to the summary.
|
76 |
+
- **Annotations**: This section includes annotations for selecting summary sentences. It is subdivided into:
|
77 |
+
- **Indexes**: Indices of sentences that have been selected for the summary. This field varies across datasets based on the number of annotators.
|
78 |
+
- **Ranking**: Rankings assigned to sentences based on their perceived importance for the summary. This feature is more prominent in datasets focusing on sentence importance ranking.
|
79 |
+
|
80 |
+
```python
|
81 |
+
{
|
82 |
+
'Title': '<title_of_document>',
|
83 |
+
'Sentences': ['<sentence_1>', '<sentence_2>', ..., '<sentence_n>'],
|
84 |
+
'Annotations': {
|
85 |
+
'Indexes': {
|
86 |
+
|
87 |
+
'Annotator1': [<index_of_selected_sentence_1>, ..., <index_of_selected_sentence_m>],
|
88 |
+
# If there are more than one annotator
|
89 |
+
'Annotator2': [...],
|
90 |
+
# etc.
|
91 |
+
},
|
92 |
+
'Ranking': {
|
93 |
+
'Annotator1': [<ranking_of_first_sentence>,<ranking_of_second_sentence>,..., <ranking_of_mth_sentence>],
|
94 |
+
# If there are more than one annotator
|
95 |
+
'Annotator2': [...],
|
96 |
+
# etc.
|
97 |
+
}
|
98 |
+
}
|
99 |
+
}
|
100 |
+
```
|
101 |
+
|
102 |
+
## Cite XTINGE Turkish Extractive Summarization Dataset
|
103 |
+
```
|
104 |
+
{
|
105 |
+
|
106 |
+
}
|
107 |
+
```
|
108 |
+
|
109 |
+
|