mrshu commited on
Commit
a252b79
·
verified ·
1 Parent(s): 4ad3312
Files changed (1) hide show
  1. README.md +64 -262
README.md CHANGED
@@ -1,298 +1,100 @@
1
  ---
 
 
 
 
2
  language:
3
  - sk
4
- pretty_name: "DrMax Pharmacy Q&A Dataset"
5
- tags:
6
- - pharmacy
7
- - healthcare
8
- - question-answering
9
- - medical
10
- - slovak
11
- - conversational
12
- license: cc-by-4.0
13
- task_categories:
14
- - question-answering
15
- - text-generation
16
- - conversational
17
  size_categories:
18
- - 1K<n<10K
 
 
 
 
 
 
19
  ---
20
 
21
- # DrMax Pharmacy Q&A Dataset
22
 
23
  ## Dataset Description
24
 
25
- This dataset contains 4,703 real-world question-answer pairs from the DrMax pharmacy website (Slovak version), scraped from the "Spýtajte sa lekárnika" (Ask the Pharmacist) section. It represents genuine healthcare and pharmaceutical questions asked by Slovak-speaking patients, answered by licensed pharmacists.
26
 
27
- ### Dataset Summary
 
28
 
29
- The dataset consists of authentic medical and pharmaceutical Q&A content covering a wide range of health topics including:
30
- - Pain management (headaches, joint pain, muscle pain)
31
- - Medication advice and drug interactions
32
- - Common ailments and symptoms
33
- - Chronic conditions (diabetes, neuropathy, arthritis)
34
- - Over-the-counter medication recommendations
35
- - General health and wellness questions
36
 
37
- Each entry contains the full question text, professional pharmacist answer, categorization tags, dates, and metadata. This makes it valuable for:
38
- - Training Slovak medical/pharmaceutical question-answering models
39
- - Healthcare chatbot development
40
- - Medical information retrieval systems
41
- - Understanding common health concerns in Slovak-speaking populations
42
- - Research in pharmaceutical and medical NLP
43
 
44
  ### Languages
45
 
46
- All content is in **Slovak (sk)**, including both patient questions and pharmacist answers.
47
-
48
- ## Dataset Structure
49
-
50
- ### Data Fields
51
-
52
- - `url` (string): The full URL to the Q&A page on DrMax.sk
53
- - `slug` (string): URL slug/identifier for the question
54
- - `title` (string): The title/summary of the question
55
- - `question` (string): Full text of the patient's question
56
- - `answer` (string): Complete pharmacist's response with recommendations
57
- - `date` (string): Publication date of the Q&A (format: DD. M. YYYY)
58
- - `tags` (list of strings): Categorization tags (average 2.04 tags per entry, 367 unique tags)
59
- - `question_author` (string): Name of the person who asked (usually empty for privacy)
60
- - `answer_author` (string): Name and credentials of the responding pharmacist
61
- - `scraped_at` (string): ISO 8601 timestamp of when the data was collected
62
-
63
- ### Data Splits
64
-
65
- This dataset contains a single split with 4,703 examples.
66
-
67
- ### Example
68
-
69
- ```python
70
- {
71
- 'url': 'https://www.drmax.sk/spytajte-sa-lekarnika/bolest-hlavy-a-nevolnost',
72
- 'slug': 'bolest-hlavy-a-nevolnost',
73
- 'title': 'bolesť hlavy a nevoľnosť',
74
- 'question': 'dobrý deň\n(Som cudzinec a môj jazyk nie je dobrý)\nMám migrénu na ľavej strane a za uchom, stav 6/10.\nSprevádzané pretrvávajúcou únavou a nevoľnosťou. Moja choroba sa stala v posledných dvoch dňoch\nĎakujem',
75
- 'answer': 'Dobrý deň,\nďakujeme za Vašu otázku. Migréna sa vyznačuje silnou a pulzujúcou bolesťou len na jednej polovici hlavy...',
76
- 'date': '25. 6. 2022',
77
- 'tags': ['bolesť', 'bolesť hlavy', 'migréna', 'tlmenie bolesti', 'nevoľnosť'],
78
- 'question_author': '',
79
- 'answer_author': 'PharmDr. Dominika Titková',
80
- 'scraped_at': '2025-11-11T17:00:33.177495'
81
- }
82
- ```
83
-
84
- ## Dataset Creation
85
-
86
- ### Source Data
87
-
88
- The Q&A pairs were collected from DrMax.sk's "Spýtajte sa lekárnika" (Ask the Pharmacist) section, where real patients ask health-related questions and licensed pharmacists provide professional answers.
89
-
90
- #### Data Collection
91
-
92
- - **Collection Date**: November 11, 2025
93
- - **Collection Method**: Automated web scraping using Scrapy and Playwright
94
- - **Source Website**: DrMax.sk (leading pharmacy chain in Slovakia and Central Europe)
95
- - **Content Type**: Public Q&A forum moderated by licensed pharmacists
96
- - **Time Range**: Questions and answers from 2022 onwards
97
-
98
- ### Annotations
99
-
100
- - **Tags**: The dataset includes 367 unique tags covering medical topics, symptoms, conditions, and treatments
101
- - **Average Tags per Entry**: 2.04
102
- - **Author Attribution**: Most answers include the pharmacist's name and credentials (PharmDr., Mgr.)
103
-
104
- ## Dataset Statistics
105
-
106
- - **Total Q&A Pairs**: 4,703
107
- - **Unique Tags**: 367
108
- - **Average Tags per Entry**: 2.04
109
- - **Questions with Tags**: 100%
110
- - **File Size**: ~2.7 MB (Parquet format)
111
- - **Language**: Slovak
112
-
113
- ### Common Topics (by tag frequency)
114
-
115
- The dataset covers a wide range of pharmaceutical and medical topics including pain management, vitamins, skin conditions, digestive issues, respiratory problems, and medication advice.
116
-
117
- ## Use Cases
118
-
119
- ### Question-Answering Systems
120
- Train or fine-tune models to answer health and pharmaceutical questions in Slovak.
121
-
122
- ### Medical Chatbots
123
- Develop conversational AI assistants for pharmacy and healthcare contexts.
124
-
125
- ### Information Retrieval
126
- Build semantic search systems for medical information in Slovak language.
127
-
128
- ### Research Applications
129
- - Study common health concerns in Slovak-speaking populations
130
- - Analyze patient-pharmacist communication patterns
131
- - Research pharmaceutical recommendations and practices
132
- - Develop medical NLP tools for Slovak language
133
-
134
- ### Educational Tools
135
- Create training materials for pharmacy students or healthcare professionals.
136
-
137
- ## Considerations for Using the Data
138
-
139
- ### Social Impact
140
-
141
- **Positive Impacts:**
142
- - Improves healthcare information accessibility in Slovak
143
- - Provides training data for medical AI systems in underrepresented language
144
- - Democratizes access to pharmaceutical knowledge
145
- - Can help reduce healthcare information gaps
146
-
147
- **Potential Risks:**
148
- - Models trained on this data should NOT replace professional medical advice
149
- - Automated systems should clearly indicate they are not substitutes for doctors/pharmacists
150
- - Misinformation risk if models generate incorrect medical advice
151
-
152
- ### Discussion of Biases
153
-
154
- **Known Biases:**
155
- - **Commercial Bias**: Answers often recommend DrMax products by name
156
- - **Demographic Bias**: Questions may reflect specific age groups or populations more than others
157
- - **Geographic Bias**: Slovak-specific pharmaceutical products and healthcare system
158
- - **Temporal Bias**: Medical knowledge and pharmaceutical recommendations from 2022+ only
159
- - **Selection Bias**: Only includes questions that were selected for publication on the website
160
-
161
- **Recommended Mitigations:**
162
- - Cross-reference with multiple medical sources
163
- - Include disclaimers about medical advice
164
- - Consider fine-tuning with more diverse medical sources
165
- - Validate recommendations against current medical guidelines
166
 
167
- ### Privacy and Ethics
168
 
169
- - **Anonymity**: Question authors are typically not named (empty field)
170
- - **Public Data**: All Q&A pairs were publicly accessible on DrMax website
171
- - **Professional Answers**: Responses provided by licensed healthcare professionals
172
- - **No Personal Health Information**: Questions are generalized and don't include identifying medical records
173
- - **Terms of Service**: Users should respect DrMax's terms when using this data
174
 
175
- **Ethical Considerations:**
176
- - This data should be used responsibly for healthcare improvement
177
- - Do not use for generating medical advice without human oversight
178
- - Consider impact on healthcare accessibility and safety
179
- - Respect patient privacy even though data is public
180
 
181
- ## Limitations
 
 
 
 
 
 
 
 
 
 
 
 
 
182
 
183
- - **Language**: Only Slovak - not transferable to other languages without translation
184
- - **Domain**: Pharmacy/OTC medication focused - not comprehensive medical knowledge
185
- - **Temporal**: Snapshot from specific time period - medical knowledge evolves
186
- - **Quality Variance**: Answer quality may vary between different pharmacists
187
- - **Product Focus**: Tendency to recommend specific branded products
188
- - **Scope**: Limited to questions deemed appropriate for pharmacist advice (no serious diagnosis)
189
 
190
- ## Additional Information
 
 
191
 
192
- ### Licensing Information
193
 
194
- This dataset is released under the **Creative Commons Attribution 4.0 International (CC BY 4.0)** license. When using this dataset, please provide appropriate attribution.
 
 
195
 
196
- **Important**: While this dataset is provided under CC BY 4.0, the original content is owned by DrMax and was sourced from their publicly accessible website. Users should:
197
- - Provide attribution to both this dataset and DrMax as the source
198
- - Not claim the medical advice as their own
199
- - Respect DrMax's terms of service
200
 
201
- **Disclaimer**: The dataset compilers are not responsible for the accuracy or completeness of the medical information. This dataset is provided "as-is" for research and educational purposes.
202
 
203
- ### Citation Information
204
-
205
- If you use this dataset, please cite:
206
-
207
- ```bibtex
208
- @dataset{drmax_pharmacy_qa_2025,
209
- title={DrMax Pharmacy Q&A Dataset},
210
- author={NaiveNeuron},
211
- year={2025},
212
- publisher={HuggingFace},
213
- url={https://huggingface.co/datasets/[your-username]/drmax-pharmacy-qa},
214
- note={4,703 Slovak pharmacy Q&A pairs from DrMax.sk}
215
- }
216
- ```
217
-
218
- ### Contributions
219
-
220
- Dataset created and maintained by the pharmacy-scrapers project.
221
-
222
- **Data Source**: DrMax.sk "Spýtajte sa lekárnika" section
223
- **Scraping Tools**: Scrapy, Playwright
224
- **Collection Date**: November 11, 2025
225
 
226
  ## Usage
227
 
228
- ### Loading the Dataset
229
-
230
- ```python
231
- from datasets import load_dataset
232
-
233
- # Load the dataset
234
- dataset = load_dataset("your-username/drmax-pharmacy-qa")
235
-
236
- # View first example
237
- print(dataset['train'][0])
238
- ```
239
-
240
- ### Example: Question-Answer Retrieval
241
-
242
- ```python
243
- from datasets import load_dataset
244
-
245
- dataset = load_dataset("your-username/drmax-pharmacy-qa")
246
-
247
- # Find all questions about headaches
248
- headache_qa = [
249
- item for item in dataset['train']
250
- if 'bolesť hlavy' in item['tags'] or 'migréna' in item['tags']
251
- ]
252
-
253
- print(f"Found {len(headache_qa)} Q&A pairs about headaches")
254
- ```
255
-
256
- ### Example: Filter by Tags
257
-
258
  ```python
259
  from datasets import load_dataset
260
 
261
- dataset = load_dataset("your-username/drmax-pharmacy-qa")
 
 
262
 
263
- # Get all unique tags
264
- all_tags = set()
265
- for item in dataset['train']:
266
- all_tags.update(item['tags'])
267
-
268
- print(f"Dataset contains {len(all_tags)} unique tags")
269
- print(f"Sample tags: {list(all_tags)[:10]}")
270
  ```
271
 
272
- ### Example: Training Data for Q&A Model
273
-
274
- ```python
275
- from datasets import load_dataset
276
-
277
- dataset = load_dataset("your-username/drmax-pharmacy-qa")
278
-
279
- # Prepare for instruction tuning
280
- def format_for_training(example):
281
- return {
282
- 'instruction': example['question'],
283
- 'response': example['answer'],
284
- 'context': f"Tags: {', '.join(example['tags'])}"
285
- }
286
-
287
- training_data = dataset['train'].map(format_for_training)
288
- ```
289
-
290
- ## Dataset Card Contact
291
-
292
- For questions, issues, or suggestions regarding this dataset:
293
- - Open an issue on the source repository
294
- - Check the documentation at the repository
295
-
296
- ---
297
 
298
- **⚠️ Medical Disclaimer**: This dataset is for research and educational purposes only. The information contained in this dataset should not be used as a substitute for professional medical advice, diagnosis, or treatment. Always seek the advice of qualified health providers with questions regarding medical conditions.
 
1
  ---
2
+ annotations_creators:
3
+ - expert-generated
4
+ language_creators:
5
+ - found
6
  language:
7
  - sk
8
+ license: other
9
+ multilinguality:
10
+ - monolingual
11
+ pretty_name: Demagog.sk Vyroky
 
 
 
 
 
 
 
 
 
12
  size_categories:
13
+ - 10K<n<100K
14
+ source_datasets:
15
+ - original
16
+ task_categories:
17
+ - text-classification
18
+ task_ids:
19
+ - fact-checking
20
  ---
21
 
22
+ # Demagog.sk Vyroky
23
 
24
  ## Dataset Description
25
 
26
+ Demagog.sk Vyroky is a collection of fact-checked political statements scraped from [Demagog.sk](https://demagog.sk/vyroky). Each record contains the original claim, the speaker, the fact-check verdict, and the supporting analysis written by the Demagog.sk editorial team.
27
 
28
+ - **Total examples:** 20495
29
+ - **Latest scrape timestamp:** 2025-10-16T21:00:22.850424+00:00
30
 
31
+ ### Supported Tasks
 
 
 
 
 
 
32
 
33
+ - **Fact-checking / claim verification:** predict the fact-check verdict (`verdict`) given the statement and optional context.
34
+ - **Evidence summarisation:** leverage the `analysis_text` to train models that generate or evaluate fact-check rationales.
35
+ - **Speaker and stance profiling:** analyse claims by political actor or party using the `speaker` and `speaker_party` fields.
 
 
 
36
 
37
  ### Languages
38
 
39
+ - Slovak (`sk`)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
40
 
41
+ ## Data Splits
42
 
43
+ | Split | Examples |
44
+ | --- | --- |
45
+ | train | 12297 |
46
+ | validation | 4099 |
47
+ | test | 4099 |
48
 
49
+ ## Data Fields
 
 
 
 
50
 
51
+ - `id`: string identifier (usually the trailing portion of the vyrok URL).
52
+ - `numeric_id`: numeric ID when available.
53
+ - `url`: canonical Demagog.sk URL for the fact-check.
54
+ - `statement`: verbatim political statement under review.
55
+ - `speaker`: full name of the speaker.
56
+ - `speaker_party`: political affiliation displayed on Demagog.sk.
57
+ - `speaker_url`: link to the speaker profile on Demagog.sk.
58
+ - `statement_date`: ISO date when the claim was made (if available).
59
+ - `verdict`: fact-check verdict label in Slovak (e.g., `Pravda`, `Nepravda`).
60
+ - `analysis_text`: editorial commentary summarising the evidence.
61
+ - `analysis_paragraphs`: list of paragraphs extracted from the commentary.
62
+ - `analysis_sources`: dictionary with `text` and `url` lists aligned per citation.
63
+ - `analysis_date`: ISO date when the analysis was published (if available).
64
+ - `scraped_at`: ISO timestamp when this dataset snapshot was collected.
65
 
66
+ ## Data Source and Collection Process
 
 
 
 
 
67
 
68
+ - Statements and annotations originate from Demagog.sk fact-check articles.
69
+ - The dataset is gathered via the public site API combined with HTML parsing of individual statement pages.
70
+ - Verdict labels and commentary are authored by Demagog.sk fact-checkers.
71
 
72
+ ## Considerations for Use
73
 
74
+ - Fact-check labels follow Demagog.sk taxonomy; users may wish to map them to English equivalents or merge classes for specific tasks.
75
+ - Commentary text is written in Slovak; downstream tasks may require translation for non-Slovak models.
76
+ - Verify licensing and usage policies of Demagog.sk before redistributing or deploying models trained on this dataset.
77
 
78
+ ## Citation
 
 
 
79
 
80
+ If you use this dataset, please cite Demagog.sk and reference this repository. An example citation:
81
 
82
+ > Demagog.sk. *Factcheck politických diskusií.* https://demagog.sk
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
83
 
84
  ## Usage
85
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
86
  ```python
87
  from datasets import load_dataset
88
 
89
+ dataset = load_dataset("NaiveNeuron/DemagogSK", name="default")
90
+ # For local files, replace the repo name with the path to this folder:
91
+ # dataset = load_dataset("path/to/demagogsk_vyroky", name="default")
92
 
93
+ train = dataset["train"]
94
+ validation = dataset["validation"]
95
+ test = dataset["test"]
 
 
 
 
96
  ```
97
 
98
+ ## License
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
99
 
100
+ The dataset inherits the terms of use of Demagog.sk. Confirm permissions for your intended use case before redistribution.