Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
csv
Languages:
Japanese
Size:
10K - 100K
ArXiv:
License:
File size: 8,660 Bytes
51b32a2 282c7fb 4781b0b 895b0c9 282c7fb 4781b0b 895b0c9 282c7fb 4781b0b 895b0c9 282c7fb 4781b0b 895b0c9 282c7fb 51b32a2 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 |
---
license: cc-by-nc-sa-4.0
task_categories:
- text-classification
language:
- ja
size_categories:
- 100K<n<1M
configs:
- config_name: ad-acceptability
default: true
data_files:
- split: train
path: "data/ad-acceptability/train.tsv"
- split: valid
path: "data/ad-acceptability/valid.tsv"
- split: test
path: "data/ad-acceptability/test.tsv"
- config_name: ad-consistency
data_files:
- split: train
path: "data/ad-consistency/train.tsv"
- split: valid
path: "data/ad-consistency/valid.tsv"
- split: test
path: "data/ad-consistency/test.tsv"
- config_name: ad-similarity
data_files:
- split: train
path: "data/ad-similarity/train.tsv"
- split: valid
path: "data/ad-similarity/valid.tsv"
- split: test
path: "data/ad-similarity/test.tsv"
- config_name: a3-recognition
data_files:
- split: train
path: "data/a3-recognition/train.tsv"
- split: valid
path: "data/a3-recognition/valid.tsv"
- split: test
path: "data/a3-recognition/test.tsv"
---
# AdTEC: A Unified Benchmark for Evaluating Text Quality in Search Engine Advertising
[](https://arxiv.org/abs/2408.05906)
[](https://github.com/cyberagentailab/adtec)
[](https://cyberagentailab.github.io/AdTEC/)
[](https://github.com/CyberAgentAILab/AdTEC/blob/main/materials/NAACL2025-poster.pdf)
[](https://github.com/CyberAgentAILab/AdTEC/blob/main/materials/NAACL2025-slides.pdf)
[](https://youtu.be/3QmKidnlkiI)
The AdTEC dataset is designed to evaluate the quality of ad texts from multiple aspects, considering practical advertising operations.
> [!NOTE]
> We are excited to announce that our paper has been accepted to the main track of [NAACL 2025](https://aclanthology.org/2025.naacl-long.391/)! We are making the data from this research publicly available in this repository.
## Experiments and Tasks Considered in the Paper
This dataset includes five tasks:
- **Ad Acceptability**: Given a text, predict the acceptance of overall quality with binary labels: `acceptable`/`unacceptable`.
- **Ad Consistency**: Given a pair of ad text and landing page (LP) text, predict the consistency between the ad and LP text with binary labels: `consistent`/`inconsistent`.
- **Ad Performance Estimation**: Given ad texts, keywords, and industry type, predict the overall performance with a score ranging from 0 to 100.
- **A3 Recognition**: Given a text, predict all possible aspects of advertising appeals (A3). The comprehensive list of A3 can be found in [[Murakami et al., 2022](https://aclanthology.org/2022.naacl-industry.9/)].
- **Ad Similarity**: Given a pair of ad texts, predict their similarity with a score ranging from 1 to 5.
## Languages
The dataset contains Japanese text only.
## Domain
Online advertisement (Search engine advertisement).
## Dataset Curators
The dataset was created by Peinan Zhang, Yusuke Sakai, Masato Mita, Hiroki Ouchi, and Taro Watanabe.
## Dataset Structure
All tasks are defined in TSV format and split into train, dev, and test sets.
The number of instances for each task in each split is shown in the table below.
| Task | Train | Dev | Test |
|------|------:|----:|-----:|
| Ad Acceptability | 13,265 | 970 | 980 |
| Ad Consistency | 10,635 | 945 | 970 |
| Ad Performance Estimation | 125,087 | 965 | 965 |
| A3 Recognition | 1,856 | 465 | 410 |
| Ad Similarity | 4,980 | 623 | 629 |
Detailed examples of each task are provided below.
### Ad Acceptability
```tsv
label title
acceptable 最新のビデオキャプチャー年間ランキングをご紹介・会員ランクに応じてポイント獲得がお得に。
unacceptable 茨城県アパート以上
acceptable 気になる保障をカンタン見積り
...
```
### Ad Consistency
```tsv
label lp_text title
consistent 介護の転職サポート登録のページ。介護の求人・転職なら介護ワーカー。介護福祉士、ホームヘルパー、ケアマネ、社会福祉士などの求人情報がどこよりも詳しくわかる！あなたの理想の職場がきっと見つかる！ 介護職の求人なら【名古屋市】
inconsistent 手ぶらで体験レッスンが0円！上下ウェア・バスタオル・フェイスタオル・ヨガマット無料レンタルだから、手ぶらでOK！さらにお水（550ml×2本）をプレゼント！ ヨガ大分スタジオ「ロイブ」
consistent ホーム コース ゲーム総合科 ゲーム総合科 プロの講師陣があなたのなりゲームクリエイターになるにはを全力サポート！資料請求はこちらから
...
```
### Ad Performance Estimation
```tsv
industry_type keyword title_1 title_2 title_3 description_1 description_2 score
EC hunter 【公式】[MASK_2]（[MASK_8]） 《MAX50% OFF》AUTUMN SALE 人気＆注目アイテムはこちらから 人気商品がMAX50%OFF！機能性抜群の[MASK_8]アイテムで秋を楽しく過ごそう 機能的でスタイリッシュなデザイン。創業当初から愛される「機能美」をあなたへ。 55.90851442
金融 かしつけ 急ぎでお金の貸付受けたい方必見 免許証だけで最短1時間貸付可能 ≪最短1h融資可能≫ローン貸付ランキング24hスマホ完結申込。融資可能か1秒診断 79.01220606
...
```
### A3 Recognition
```tsv
title labels
THE THOUSAND KYOTO（ザ・サウザンド京都）の最寄駅は京都駅。THE THOUSAND KYOTO。ブライダルフェア情報、おトクな割引特典、フォトギャラリー、口コミ、見積もり例が充実！結婚式場探しはハナユメ！ 商品/サービス特徴|オファー|アクセス
鳥取ガス enetopia
ジャニーズ屈指のダンス力を誇る7人組グループ「Travis Japan」を全4週にわたって特集！ 商品/サービス特徴
...
```
### Ad Similarity
```tsv
text1 text2 score
３月３１日まで実施中 山口トヨタ／お客様大感謝祭 2.67
春得キャンペーン第1弾 春得キャンペーン 4.33
...
```
## License
AdTEC dataset is released under the [CreativeCommons Attribution-NonCommercial-ShareAlike 4.0 International license](./LICENSE).
## Citation
```latex
@inproceedings{zhang2025adtec,
title={{AdTEC}: A Unified Benchmark for Evaluating Text Quality in Search Engine Advertising},
author={Peinan Zhang and Yusuke Sakai and Masato Mita and Hiroki Ouchi and Taro Watanabe},
booktitle={Proceedings of the 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL)},
year={2025},
publisher={Association for Computational Linguistics},
eprint={2408.05906},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2408.05906},
}
```
|