siimh commited on
Commit
b2839ef
·
1 Parent(s): d88ea48

Add a cleaned version of the original dataset, with morphological tags removed, leaving only the plain text

Browse files
Files changed (3) hide show
  1. .gitattributes +2 -0
  2. README.md +148 -0
  3. corpus_et_clean.jsonl +3 -0
.gitattributes CHANGED
@@ -56,3 +56,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
56
  # Video files - compressed
57
  *.mp4 filter=lfs diff=lfs merge=lfs -text
58
  *.webm filter=lfs diff=lfs merge=lfs -text
 
 
 
56
  # Video files - compressed
57
  *.mp4 filter=lfs diff=lfs merge=lfs -text
58
  *.webm filter=lfs diff=lfs merge=lfs -text
59
+ corpus_et.jsonl filter=lfs diff=lfs merge=lfs -text
60
+ corpus_et_clean.jsonl filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,148 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Estonian National Corpus 2021: Morphologically Tagged and Clean Text
2
+
3
+ ## Dataset Summary
4
+
5
+ This repository contains two versions of the **Estonian National Corpus 2021**, representing around **43GB of data** and **millions of sentences** annotated with **morphological features**. This corpus is designed to be useful for:
6
+ - **Morphological analysis**.
7
+ - **Natural language understanding** for **Estonian**.
8
+
9
+ Both files in this dataset originate from the **Estonian National Corpus (ENC) 2021** and have been processed for different use cases:
10
+ 1. **`corpus_et.jsonl`**: Contains **morphologically tagged text** where each word is annotated with its base lemma and morphological features (e.g., case, number, tense).
11
+ 2. **`corpus_et_clean.jsonl`**: A **cleaned version** of the original dataset, with **morphological tags removed**, leaving only the plain text.
12
+
13
+ These datasets can be used for fine-tuning **language models** or studying **morphological patterns** and **syntactic structures** in the Estonian language.
14
+
15
+ ---
16
+
17
+ ## Files in This Dataset:
18
+ - **`corpus_et.jsonl`**: This file contains text with **morphological annotations**, where each word is followed by a tag indicating its grammatical properties (e.g., `Vabadus_vabadus-s`).
19
+
20
+ The fields in this file are as follows:
21
+ - **`text`** (string): A sentence where each word is attached to its morphological tag. Example:
22
+ ```json
23
+ {"text": "Vabadus_vabadus-s millest_mis-p ilmselt_ilmselt-d ei_ei-v."}
24
+ ```
25
+
26
+ - **`corpus_et_clean.jsonl`**: This file contains the **same sentences** as `corpus_et.jsonl` but with the **morphological tags stripped**. It is clean text suitable for general **language fine-tuning**, without the added complexity of morph tags.
27
+
28
+ The fields in this file are as follows:
29
+ - **`text`** (string): A plain-text sentence, with the same meaning as the original tagged version. Example:
30
+ ```json
31
+ {"text": "Vabadus millest ilmselt ei."}
32
+ ```
33
+
34
+ ---
35
+
36
+ ## Dataset Structure
37
+
38
+ - Both datasets are in JSONL (newline-delimited JSON) format.
39
+ - The datasets do not contain explicit **train/validation splits**, meaning users can split them according to their needs.
40
+ - Each line is a JSON object consisting of the `"text"` field for the sentence.
41
+
42
+ ---
43
+
44
+ ## Use Cases
45
+ This dataset is highly versatile and could be used in multiple **natural language processing (NLP) tasks**, such as:
46
+ 1. **Tagging and Morphological Parsing**: Use the **tagged version** (`corpus_et.jsonl`) to train models that focus on **morphological understanding** or performing **syntax parsing** for Estonian.
47
+ 2. **General Language Modeling**: Use the **cleaned version** (`corpus_et_clean.jsonl`) to train **general-purpose Estonian language models** for tasks like **text generation** or **machine translation**.
48
+ 3. **Inflection Modeling**: The tagged data can also serve as a training set for **inflection generation models** that learn to correctly inflect Estonian words based on grammatical context.
49
+
50
+ ---
51
+
52
+ ## Dataset Statistics
53
+
54
+ The dataset has the following statistics (for both `corpus_et.jsonl` and `corpus_et_clean.jsonl`):
55
+
56
+ - **Total size**: ~43GB
57
+ - **Sentences**: Approx. **196 million** sentences
58
+ - **Words**: Approx. **2.4 billion** words
59
+ - **Documents**: Approx. **11.7 million** documents
60
+ - **Paragraphs**: Approx. **64.5 million** paragraphs
61
+
62
+ ---
63
+
64
+ ## Examples
65
+
66
+ ### Example from `corpus_et.jsonl` (Tagged Text):
67
+ ```json
68
+ {
69
+ "text": "Vabadus_vabadus-s millest_mis-p ilmselt_ilmselt-d ei_ei-v."
70
+ }
71
+ ```
72
+
73
+ ### Example from `corpus_et_clean.jsonl` (Clean Text):
74
+ ```json
75
+ {
76
+ "text": "Vabadus millest ilmselt ei."
77
+ }
78
+ ```
79
+
80
+ ---
81
+
82
+ ## Dataset Creation
83
+
84
+ The dataset was derived from the **Estonian National Corpus 2021** and processed into these two versions:
85
+ 1. **Tagged JSONL** (`corpus_et.jsonl`): This version contains **morphologically annotated text** based on the original corpus.
86
+ 2. **Clean JSONL** (`corpus_et_clean.jsonl`): This version contains the same text but without the **morphological tags**—resulting in clean, plain-text sentences.
87
+
88
+ ---
89
+
90
+ ## License
91
+
92
+ This dataset is made available under the **Creative Commons Attribution License (CC-BY)**, allowing free use and sharing with proper attribution to the authors listed below.
93
+
94
+ ---
95
+
96
+ ## Citation
97
+
98
+ If you use this dataset, please cite the authors of the original **Estonian National Corpus 2021**:
99
+ ```plaintext
100
+ Koppel, Kristina; Kallas, Jelena (2022). Eesti keele ühendkorpus 2021. DOI: 10.15155/3-00-0000-0000-0000-08E60L
101
+ ```
102
+
103
+ Please also give credit to this dataset repository when used in a project:
104
+ ```plaintext
105
+ @dataset{siimh_estonian_corpus_2021,
106
+ author = {Siim Haugas},
107
+ title = {Estonian National Corpus 2021: Tagged and Cleaned Versions},
108
+ year = {2023},
109
+ publisher = {Hugging Face},
110
+ url = {https://huggingface.co/datasets/siimh/estonian_corpus_2021}
111
+ }
112
+ ```
113
+
114
+ ---
115
+
116
+ ## Contact
117
+
118
+ For any inquiries regarding this dataset or data processing, feel free to contact:
119
+ - **Siim Haugas**
120
+ - Hugging Face profile: [https://huggingface.co/siimh](https://huggingface.co/siimh)
121
+
122
+ ---
123
+
124
+ ## How to Load the Dataset
125
+
126
+ You can load the dataset directly from Hugging Face with the **datasets** library.
127
+
128
+ ### For the Tagged Version (`corpus_et.jsonl`):
129
+ ```python
130
+ from datasets import load_dataset
131
+
132
+ # Load the tagged dataset
133
+ dataset = load_dataset("siimh/estonian_corpus_2021", data_files="corpus_et.jsonl")
134
+ ```
135
+
136
+ ### For the Cleaned Version (`corpus_et_clean.jsonl`):
137
+ ```python
138
+ from datasets import load_dataset
139
+
140
+ # Load the clean dataset
141
+ dataset = load_dataset("siimh/estonian_corpus_2021", data_files="corpus_et_clean.jsonl")
142
+ ```
143
+
144
+ ---
145
+
146
+ ### Conclusion
147
+ This dataset is a powerful resource for fine-tuning Estonian-language models, improving **morphological awareness**, and supporting natural language processing tasks. Whether you're training a **language model**, working on **syntactic parsing**, or preparing for a Estonian **chatbot**, this dataset will provide excellent material for various NLP tasks.
148
+
corpus_et_clean.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f1120cb8793e8489ba0240a5c31c4a504694757180c35ca01b6e0eef0c53ffd4
3
+ size 20710737846