Update README.md
#3
by
chris-at-bonafide
- opened
README.md
CHANGED
@@ -1,8 +1,30 @@
|
|
1 |
|
2 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
|
4 |
## Dataset Description
|
5 |
|
|
|
|
|
|
|
|
|
|
|
6 |
The dataset consists of the following columns:
|
7 |
|
8 |
- `bonafide_id`: A unique identifier for the property.
|
@@ -16,7 +38,8 @@ The dataset consists of the following columns:
|
|
16 |
- `prompt`: LLM prompt regarding the presence of the feature.
|
17 |
- `expected_answer`: The expected answer (Yes/No) to the prompt.
|
18 |
|
19 |
-
|
|
|
20 |
|
21 |
| bonafide_id | property_name | address | city | state | country | feature_name | feature_type | prompt | expected_answer |
|
22 |
| ------------------------- | ----------------------------------- | -------------------- | -------- | ----- | ------- | --------------------------------------- | ---------------------------- | -------------------------------------------------------------------------------------- | --------------- |
|
@@ -25,13 +48,24 @@ The dataset consists of the following columns:
|
|
25 |
| HO-00005-US-AZ-D0E19AFB-8 | residence inn phoenix west avondale | 11410 west hotel way | Avondale | AZ | US | Meeting rooms and conference facilities | Meeting and Event Spaces | Does residence inn phoenix west avondale have Meeting rooms and conference facilities? | Yes |
|
26 |
| HO-00005-US-AZ-D0E19AFB-8 | residence inn phoenix west avondale | 11410 west hotel way | Avondale | AZ | US | Laundry and dry-cleaning services | Additional Services | Does residence inn phoenix west avondale have Laundry and dry-cleaning services? | Yes |
|
27 |
| HO-00005-US-AZ-D0E19AFB-8 | residence inn phoenix west avondale | 11410 west hotel way | Avondale | AZ | US | 24-hour front desk | Reception and Guest Services | Does residence inn phoenix west avondale have a 24-hour front desk? | Yes |
|
28 |
-
|
|
|
29 |
|
30 |
This dataset can be used for various tasks such as:
|
31 |
|
32 |
-
-
|
|
|
33 |
- Natural language processing (NLP) tasks related to question-answering systems.
|
34 |
-
- Creating or
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
35 |
|
36 |
## License
|
37 |
|
@@ -39,7 +73,7 @@ This dataset is licensed under the Creative Commons Attribution 4.0 Internationa
|
|
39 |
|
40 |
## Citation
|
41 |
|
42 |
-
bonafide.ai 2024.
|
43 |
|
44 |
## Contact
|
45 |
|
|
|
1 |
|
2 |
+
## Bonafide.ai *Hospitality* Dataset
|
3 |
+
Welcome to the Bonafide.ai Hospitality Dataset! This dataset is part of the
|
4 |
+
[Bonafide.ai](https://Bonafide.ai) *Interrogator* project, which evaluates a
|
5 |
+
select group of public-facing LLM models, derivative travel apps, and LLM
|
6 |
+
agents on their ability to accurately answer simple ternary questions about
|
7 |
+
commonly available products and services. This data set focuses on travel
|
8 |
+
hospitality (hotels, motels, lodgings, etc.)
|
9 |
+
|
10 |
+
## Project Overview
|
11 |
+
The *Interrogator* project aims to provide model and application developers,
|
12 |
+
users, and commercial enterprises with insights into how products and
|
13 |
+
services are represented int the generative AI ecosystem. The intent of
|
14 |
+
providing this data set here is to allow model and app developers to have
|
15 |
+
an easy to use evaluation data set for understanding the accuracy of there
|
16 |
+
model development efforts. [Bonafide.ai](https://Bonafide.ai) is providing
|
17 |
+
this vendor-validated data, so that consumers receive accurate and reliable
|
18 |
+
information about the goods and services they are inquiring about.
|
19 |
+
|
20 |
|
21 |
## Dataset Description
|
22 |
|
23 |
+
This dataset contains evaluation data collected directly from
|
24 |
+
service providers or, in some cases, from industry-standard sources. Our
|
25 |
+
goal is to make this data as accurate as possible through rigorous field
|
26 |
+
collection methods. However, occasional inaccuracies may occur.
|
27 |
+
|
28 |
The dataset consists of the following columns:
|
29 |
|
30 |
- `bonafide_id`: A unique identifier for the property.
|
|
|
38 |
- `prompt`: LLM prompt regarding the presence of the feature.
|
39 |
- `expected_answer`: The expected answer (Yes/No) to the prompt.
|
40 |
|
41 |
+
|
42 |
+
### Example Rows
|
43 |
|
44 |
| bonafide_id | property_name | address | city | state | country | feature_name | feature_type | prompt | expected_answer |
|
45 |
| ------------------------- | ----------------------------------- | -------------------- | -------- | ----- | ------- | --------------------------------------- | ---------------------------- | -------------------------------------------------------------------------------------- | --------------- |
|
|
|
48 |
| HO-00005-US-AZ-D0E19AFB-8 | residence inn phoenix west avondale | 11410 west hotel way | Avondale | AZ | US | Meeting rooms and conference facilities | Meeting and Event Spaces | Does residence inn phoenix west avondale have Meeting rooms and conference facilities? | Yes |
|
49 |
| HO-00005-US-AZ-D0E19AFB-8 | residence inn phoenix west avondale | 11410 west hotel way | Avondale | AZ | US | Laundry and dry-cleaning services | Additional Services | Does residence inn phoenix west avondale have Laundry and dry-cleaning services? | Yes |
|
50 |
| HO-00005-US-AZ-D0E19AFB-8 | residence inn phoenix west avondale | 11410 west hotel way | Avondale | AZ | US | 24-hour front desk | Reception and Guest Services | Does residence inn phoenix west avondale have a 24-hour front desk? | Yes |
|
51 |
+
|
52 |
+
### Usage
|
53 |
|
54 |
This dataset can be used for various tasks such as:
|
55 |
|
56 |
+
- Evaluating LLMs, derivative applications, and agents for providing accurate information about travel hospitality locations.
|
57 |
+
- Training machine learning models to recognize and verify property features/amenities.
|
58 |
- Natural language processing (NLP) tasks related to question-answering systems.
|
59 |
+
- Creating, improving, or validating a service provider's property management system(s).
|
60 |
+
|
61 |
+
|
62 |
+
## Purpose
|
63 |
+
The primary objective of the Bonafide.ai Hospitality Dataset is to ensure
|
64 |
+
that all components of the generative AI ecosystem provide accurate
|
65 |
+
information. This helps build consumer confidence in the information provided through generative AI technologies.
|
66 |
+
|
67 |
+
For more information, please visit Bonafide.ai.
|
68 |
+
|
69 |
|
70 |
## License
|
71 |
|
|
|
73 |
|
74 |
## Citation
|
75 |
|
76 |
+
[Bonafide.ai](https://bonafide.ai) (c) 2024. *Hospitality Dataset**.
|
77 |
|
78 |
## Contact
|
79 |
|