Datasets:
Tasks:
Text Generation
Modalities:
Text
Formats:
parquet
Languages:
Italian
Size:
10M - 100M
License:
Update README.md
Browse files
README.md
CHANGED
@@ -24,7 +24,7 @@ tags:
|
|
24 |
- medical
|
25 |
- biology
|
26 |
size_categories:
|
27 |
-
-
|
28 |
---
|
29 |
|
30 |
From this repository you can download the **BioBERT_Italian** dataset.
|
@@ -33,6 +33,16 @@ From this repository you can download the **BioBERT_Italian** dataset.
|
|
33 |
|
34 |
Due to the unavailability of an Italian equivalent for the millions of abstracts and full-text scientific papers used by English, BERT-based biomedical models, we leveraged machine translation to obtain an Italian biomedical corpus based on PubMed abstracts and train [**BioBIT**](https://www.sciencedirect.com/science/article/pii/S1532046423001521).
|
35 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
36 |
**BioBIT Model**
|
37 |
|
38 |
[**BioBIT**](https://www.sciencedirect.com/science/article/pii/S1532046423001521) has been evaluated on 3 downstream tasks: **NER** (Named Entity Recognition), extractive **QA** (Question Answering), **RE** (Relation Extraction).
|
@@ -54,7 +64,7 @@ Here are the results, summarized:
|
|
54 |
|
55 |
**MedPsyNIT Model**
|
56 |
|
57 |
-
We also [**fine-tuned BioBIT**](https://www.sciencedirect.com/science/article/pii/S1532046423002782) on [**PsyNIT**](IVN-RIN/PsyNIT) (Psychiatric Ner for ITalian), a native Italian **NER** (Named Entity Recognition) dataset, composed by [Italian Research Hospital Centro San Giovanni Di Dio Fatebenefratelli](https://www.fatebenefratelli.it/strutture/irccs-brescia).
|
58 |
|
59 |
**Correspondence to**
|
60 |
|
|
|
24 |
- medical
|
25 |
- biology
|
26 |
size_categories:
|
27 |
+
- 1B<n<10B
|
28 |
---
|
29 |
|
30 |
From this repository you can download the **BioBERT_Italian** dataset.
|
|
|
33 |
|
34 |
Due to the unavailability of an Italian equivalent for the millions of abstracts and full-text scientific papers used by English, BERT-based biomedical models, we leveraged machine translation to obtain an Italian biomedical corpus based on PubMed abstracts and train [**BioBIT**](https://www.sciencedirect.com/science/article/pii/S1532046423001521).
|
35 |
|
36 |
+
Corpus statistics:
|
37 |
+
- Total Tokens*: 6.2 B
|
38 |
+
- Average tokens per example: 359
|
39 |
+
- Max tokens per example: 2132
|
40 |
+
- Min tokens per example: 5
|
41 |
+
- Standard deviation: 137
|
42 |
+
|
43 |
+
*Tokenization with [**BioBIT**](https://huggingface.co/IVN-RIN/bioBIT) tokenizer
|
44 |
+
|
45 |
+
|
46 |
**BioBIT Model**
|
47 |
|
48 |
[**BioBIT**](https://www.sciencedirect.com/science/article/pii/S1532046423001521) has been evaluated on 3 downstream tasks: **NER** (Named Entity Recognition), extractive **QA** (Question Answering), **RE** (Relation Extraction).
|
|
|
64 |
|
65 |
**MedPsyNIT Model**
|
66 |
|
67 |
+
We also [**fine-tuned BioBIT**](https://www.sciencedirect.com/science/article/pii/S1532046423002782) on [**PsyNIT**](https://huggingface.co/IVN-RIN/PsyNIT) (Psychiatric Ner for ITalian), a native Italian **NER** (Named Entity Recognition) dataset, composed by [Italian Research Hospital Centro San Giovanni Di Dio Fatebenefratelli](https://www.fatebenefratelli.it/strutture/irccs-brescia).
|
68 |
|
69 |
**Correspondence to**
|
70 |
|