Upload README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,53 @@
|
|
| 1 |
-
---
|
| 2 |
-
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
- nl
|
| 5 |
+
- de
|
| 6 |
+
- fr
|
| 7 |
+
- it
|
| 8 |
+
- es
|
| 9 |
+
|
| 10 |
+
license: mit
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
# bert-base-multilingual-uncased-sentiment
|
| 14 |
+
|
| 15 |
+
This a bert-base-multilingual-uncased model finetuned for sentiment analysis on product reviews in six languages: English, Dutch, German, French, Spanish and Italian. It predicts the sentiment of the review as a number of stars (between 1 and 5).
|
| 16 |
+
|
| 17 |
+
This model is intended for direct use as a sentiment analysis model for product reviews in any of the six languages above, or for further finetuning on related sentiment analysis tasks.
|
| 18 |
+
|
| 19 |
+
## Training data
|
| 20 |
+
|
| 21 |
+
Here is the number of product reviews we used for finetuning the model:
|
| 22 |
+
|
| 23 |
+
| Language | Number of reviews |
|
| 24 |
+
| -------- | ----------------- |
|
| 25 |
+
| English | 150k |
|
| 26 |
+
| Dutch | 80k |
|
| 27 |
+
| German | 137k |
|
| 28 |
+
| French | 140k |
|
| 29 |
+
| Italian | 72k |
|
| 30 |
+
| Spanish | 50k |
|
| 31 |
+
|
| 32 |
+
## Accuracy
|
| 33 |
+
|
| 34 |
+
The finetuned model obtained the following accuracy on 5,000 held-out product reviews in each of the languages:
|
| 35 |
+
|
| 36 |
+
- Accuracy (exact) is the exact match on the number of stars.
|
| 37 |
+
- Accuracy (off-by-1) is the percentage of reviews where the number of stars the model predicts differs by a maximum of 1 from the number given by the human reviewer.
|
| 38 |
+
|
| 39 |
+
|
| 40 |
+
| Language | Accuracy (exact) | Accuracy (off-by-1) |
|
| 41 |
+
| -------- | ---------------------- | ------------------- |
|
| 42 |
+
| English | 67% | 95%
|
| 43 |
+
| Dutch | 57% | 93%
|
| 44 |
+
| German | 61% | 94%
|
| 45 |
+
| French | 59% | 94%
|
| 46 |
+
| Italian | 59% | 95%
|
| 47 |
+
| Spanish | 58% | 95%
|
| 48 |
+
|
| 49 |
+
## Contact
|
| 50 |
+
|
| 51 |
+
In addition to this model, [NLP Town](https://www.nlp.town) offers custom, monolingual sentiment models for many languages and an improved multilingual model through [RapidAPI](https://rapidapi.com/nlp-town-nlp-town-default/api/multilingual-sentiment-analysis2/).
|
| 52 |
+
|
| 53 |
+
Feel free to contact us for questions, feedback and/or requests for similar models.
|