Datasets:

Modalities:
Text
Formats:
json
Languages:
English
ArXiv:
Libraries:
Datasets
pandas
License:
jchevallard commited on
Commit
081092b
·
1 Parent(s): 873660c

Update README

Browse files
Files changed (1) hide show
  1. README.md +28 -1
README.md CHANGED
@@ -32,4 +32,31 @@ For an easier handling and download of the dataset, we have split the 2706 rows
32
  - `invalid` for any answer == "invalid question"
33
  - `no_answer` for any answer == "i don't know"
34
  - `valid` for any other answer
35
- 2. We have considered the labels `answer_type`, `domain`, `question_type` and `static_or_dynamic` and performed stratified sampling, splitting the datasets in 5 subsamples. Each subsample has thus the same statistical properties of the full dataset.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
32
  - `invalid` for any answer == "invalid question"
33
  - `no_answer` for any answer == "i don't know"
34
  - `valid` for any other answer
35
+ 2. We have considered the labels `answer_type`, `domain`, `question_type` and `static_or_dynamic` and performed stratified sampling, splitting the datasets in 5 subsamples. Each subsample has thus the same statistical properties of the full dataset.
36
+
37
+ We report below the data schema as provided in [CRAG's GitHub repository](https://github.com/facebookresearch/CRAG/blob/main/docs/dataset.md).
38
+
39
+ ## Data Schema
40
+
41
+ | Field Name | Type | Description |
42
+ |------------------------|---------------|-----------------------------------------------------------------------------------------------------------------------------------------------------------------------|
43
+ | `interaction_id` | string | A unique identifier for each example. |
44
+ | `query_time` | string | Date and time when the query and the web search occurred. |
45
+ | `domain` | string | Domain label for the query. Possible values: "finance", "music", "movie", "sports", "open". "Open" includes any factual queries not among the previous four domains. |
46
+ | `question_type` | string | Type label about the query. Possible values include: "simple", "simple_w_condition", "comparison", "aggregation", "set", "false_premise", "post-processing", "multi-hop". |
47
+ | `static_or_dynamic` | string | Indicates whether the answer to a question changes and the expected rate of change. Possible values: "static", "slow-changing", "fast-changing", and "real-time". |
48
+ | `query` | string | The question for RAG to answer. |
49
+ | `answer` | string | The gold standard answer to the question. |
50
+ | `alt_ans` | list | Other valid gold standard answers to the question. |
51
+ | `split` | integer | Data split indicator, where 0 is for validation and 1 is for the public test. |
52
+ | `search_results` | list of JSON | Contains up to `k` HTML pages for each query (`k=5` for Task #1 and `k=50` for Task #3), including page name, URL, snippet, full HTML, and last modified time. |
53
+
54
+ ### Search Results Detail
55
+
56
+ | Key | Type | Description |
57
+ |----------------------|--------|---------------------------------------------------------|
58
+ | `page_name` | string | The name of the webpage. |
59
+ | `page_url` | string | The URL of the webpage. |
60
+ | `page_snippet` | string | A short paragraph describing the major content of the page. |
61
+ | `page_result` | string | The full HTML of the webpage. |
62
+ | `page_last_modified` | string | The time when the page was last modified. |