Published 2020
| Version v5
Advertisement CTR Prediction Data
Creators
Description
Description
The Advertisement CTR Prediction dataset is a large‑scale, anonymized collection of user‑level advertising behavior recorded over seven consecutive days, split into a training set and a held‑out test set for click‑through‑rate (CTR) modeling. Each record includes timestamps, ad identifiers, user and device features, impression counts, and binary click outcomes, enabling researchers to develop and benchmark high‑accuracy CTR prediction algorithms for precision marketing. The data were released as part of a public competition aimed at advancing state‑of‑the‑art advertising models.
Variables
| Name | Description |
|---|---|
| label | Label |
| uid | Unique user ID after data anonymization |
| task_id | Unique ID of an ad task |
| adv_id | Unique ID of an ad material |
| creat_type_cd | Unique ID of an ad creative type |
| adv_prim_id | Advertiser ID of an ad task |
| dev_id | Developer ID of an ad task |
| inter_typ_cd | Display form of an ad material |
| slot_id | Ad slot ID |
| spread_app_id | App ID of an ad task |
| tags | App tag of an ad task |
| app_first_class | App level-1 category of an ad task |
| app_second_class | App level-2 category of an ad task |
| age | User age |
| city | Resident city of a user |
| city_rank | Level of the resident city of a user |
| device_name | Phone model used by a user |
| device_size | Size of the phone used by a user |
| career | User occupation |
| gender | User gender |
| net_type | Network status when a behavior occurs |
| residence | Resident province of a user |
| his_app_size | App storage size |
| his_on_shelf_time | Release time |
| app_score | App rating score |
| emui_dev | EMUI version |
| list_time | Model release time |
| device_price | Device price |
| up_life_duration | HUAWEI ID lifecycle |
| up_membership_grade | Service membership level |
| membership_life_duration | Membership lifecycle |
| consume_purchase | Paid user tag |
| communication_onlinerate | Active time by mobile phone |
| communication_avgonline_30d | Daily active time by mobile phone |
| indu_name | Ad industry information |
| pt_d | Date when a behavior occurs |
Details
| Resource type | Open dataset |
| Title | Advertisement CTR Prediction Data |
| Creators |
|
| Publisher | Kaggle |
| Year of publication | 2020 |
| Research fields | Business Administration Psychology Economic & Social History Communication Sciences Educational Research Other |
| Size | 6.86 GB |
| Formats | Comma-separated values (CSV) (.csv) |
| External resource | https://www.kaggle.com/louischen7/2020-digix-advertisement-ctr-prediction |
Additional details
Related works
- Is cited by
- Journal article: https://www.researchgate.net/publication/355243680_A_Novel_Integrated_Network_with_LightGBM_for_Click-Through_Rate_Prediction (URL)