Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
@@ -19,16 +19,16 @@ Wisesight Sentiment Corpus: Social media messages in Thai language with sentimen
|
|
19 |
* Time period: Around 2016 to early 2019. With small amount from other period.
|
20 |
* Domains: Mixed. Majority are consumer products and services (restaurants, cosmetics, drinks, car, hotels), with some current affairs.
|
21 |
* Privacy:
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
* Alternations and modifications:
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
* More characteristics of the data can be explore: https://github.com/PyThaiNLP/wisesight-sentiment/blob/master/exploration.ipynb
|
33 |
|
34 |
|
@@ -39,25 +39,25 @@ tha
|
|
39 |
## Supported Tasks
|
40 |
|
41 |
Sentiment Analysis
|
42 |
-
|
43 |
## Dataset Usage
|
44 |
### Using `datasets` library
|
45 |
```
|
46 |
-
|
47 |
-
|
48 |
```
|
49 |
### Using `seacrowd` library
|
50 |
```import seacrowd as sc
|
51 |
# Load the dataset using the default config
|
52 |
-
|
53 |
# Check all available subsets (config names) of the dataset
|
54 |
-
|
55 |
# Load the dataset using a specific config
|
56 |
-
|
57 |
```
|
58 |
-
|
59 |
-
|
60 |
-
|
61 |
|
62 |
## Dataset Homepage
|
63 |
|
|
|
19 |
* Time period: Around 2016 to early 2019. With small amount from other period.
|
20 |
* Domains: Mixed. Majority are consumer products and services (restaurants, cosmetics, drinks, car, hotels), with some current affairs.
|
21 |
* Privacy:
|
22 |
+
* Only messages that made available to the public on the internet (websites, blogs, social network sites).
|
23 |
+
* For Facebook, this means the public comments (everyone can see) that made on a public page.
|
24 |
+
* Private/protected messages and messages in groups, chat, and inbox are not included.
|
25 |
* Alternations and modifications:
|
26 |
+
* Keep in mind that this corpus does not statistically represent anything in the language register.
|
27 |
+
* Large amount of messages are not in their original form. Personal data are removed or masked.
|
28 |
+
* Duplicated, leading, and trailing whitespaces are removed. Other punctuations, symbols, and emojis are kept intact.
|
29 |
+
(Mis)spellings are kept intact.
|
30 |
+
* Messages longer than 2,000 characters are removed.
|
31 |
+
* Long non-Thai messages are removed. Duplicated message (exact match) are removed.
|
32 |
* More characteristics of the data can be explore: https://github.com/PyThaiNLP/wisesight-sentiment/blob/master/exploration.ipynb
|
33 |
|
34 |
|
|
|
39 |
## Supported Tasks
|
40 |
|
41 |
Sentiment Analysis
|
42 |
+
|
43 |
## Dataset Usage
|
44 |
### Using `datasets` library
|
45 |
```
|
46 |
+
from datasets import load_dataset
|
47 |
+
dset = datasets.load_dataset("SEACrowd/wisesight_thai_sentiment", trust_remote_code=True)
|
48 |
```
|
49 |
### Using `seacrowd` library
|
50 |
```import seacrowd as sc
|
51 |
# Load the dataset using the default config
|
52 |
+
dset = sc.load_dataset("wisesight_thai_sentiment", schema="seacrowd")
|
53 |
# Check all available subsets (config names) of the dataset
|
54 |
+
print(sc.available_config_names("wisesight_thai_sentiment"))
|
55 |
# Load the dataset using a specific config
|
56 |
+
dset = sc.load_dataset_by_config_name(config_name="<config_name>")
|
57 |
```
|
58 |
+
|
59 |
+
More details on how to load the `seacrowd` library can be found [here](https://github.com/SEACrowd/seacrowd-datahub?tab=readme-ov-file#how-to-use).
|
60 |
+
|
61 |
|
62 |
## Dataset Homepage
|
63 |
|