Datasets:

Modalities:
Text
Formats:
csv
Libraries:
Datasets
pandas
License:
jannikskytt commited on
Commit
b8287de
·
verified ·
1 Parent(s): 1050884

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +15 -19
README.md CHANGED
@@ -62,31 +62,27 @@ configs:
62
 
63
  ---
64
 
65
- # Danish medical word embeddings
66
 
67
- MeDa-We was trained on a Danish medical corpus of 123M tokens. The word embeddings are 300-dimensional and are trained using [FastText](https://fasttext.cc/).
68
-
69
- The embeddings were trained for 10 epochs using a window size of 5 and 10 negative samples.
70
-
71
- The development of the corpus and word embeddings is described further in our [paper](https://aclanthology.org/2023.nodalida-1.31/).
72
-
73
- We also trained a transformer model on the developed corpus which can be found [here](https://huggingface.co/jannikskytt/MeDa-Bert).
74
 
75
  ### Citing
76
 
77
  ```
78
- @inproceedings{pedersen-etal-2023-meda,
79
- title = "{M}e{D}a-{BERT}: A medical {D}anish pretrained transformer model",
80
- author = "Pedersen, Jannik and
81
- Laursen, Martin and
82
- Vinholt, Pernille and
 
83
  Savarimuthu, Thiusius Rajeeth",
84
- booktitle = "Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa)",
85
- month = may,
86
  year = "2023",
87
- address = "T{\'o}rshavn, Faroe Islands",
88
- publisher = "University of Tartu Library",
89
- url = "https://aclanthology.org/2023.nodalida-1.31",
90
- pages = "301--307",
 
91
  }
92
  ```
 
62
 
63
  ---
64
 
65
+ # Danish medical word embedding evaluation
66
 
67
+ The development of the datasets is described further in our [paper](https://aclanthology.org/2023.nejlt-1.4/).
 
 
 
 
 
 
68
 
69
  ### Citing
70
 
71
  ```
72
+ @inproceedings{laursen-etal-2023-benchmark,
73
+ title = "Benchmark for Evaluation of {D}anish Clinical Word Embeddings",
74
+ author = "Laursen, Martin Sundahl and
75
+ Pedersen, Jannik Skyttegaard and
76
+ Vinholt, Pernille Just and
77
+ Hansen, Rasmus S{\o}gaard and
78
  Savarimuthu, Thiusius Rajeeth",
79
+ editor = "Derczynski, Leon",
80
+ booktitle = "Northern European Journal of Language Technology, Volume 9",
81
  year = "2023",
82
+ address = {Link{\"o}ping, Sweden},
83
+ publisher = {Link{\"o}ping University Electronic Press},
84
+ url = "https://aclanthology.org/2023.nejlt-1.4",
85
+ doi = "https://doi.org/10.3384/nejlt.2000-1533.2023.4132",
86
+ abstract = "In natural language processing, benchmarks are used to track progress and identify useful models. Currently, no benchmark for Danish clinical word embeddings exists. This paper describes the development of a Danish benchmark for clinical word embeddings. The clinical benchmark consists of ten datasets: eight intrinsic and two extrinsic. Moreover, we evaluate word embeddings trained on text from the clinical domain, general practitioner domain and general domain on the established benchmark. All the intrinsic tasks of the benchmark are publicly available.",
87
  }
88
  ```