Summary of Transferable and Forecastable User Targeting Foundation Model, by Bin Dou et al.
Transferable and Forecastable User Targeting Foundation Model
by Bin Dou, Baokun Wang, Yun Zhu, Xiaotong Lin, Yike Xu, Xiaorui Huang, Yang Chen, Yun Liu, Shaoshuai Han, Yongchao Liu, Tianyi Zhang, Yu Cheng, Weiqiang Wang, Chuntao Hong
First submitted to arxiv on: 17 Dec 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed FOUND model for user targeting addresses two significant challenges: poor transferability and insufficient forecastability in real-world applications. The framework integrates multi-scenario user data, aligning it with one-sentence targeting demand inputs through contrastive pre-training. This improves cross-domain transferability. For improved forecastability, the text description of each user is derived based on anticipated future behaviors, while user representations are constructed from historical information. Experimental results show that FOUND outperforms existing baselines in cross-domain, real-world user targeting scenarios. The model has been successfully deployed on the Alipay platform and widely utilized across various scenarios. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary User targeting helps marketers choose the right people for their ads. Current methods have two big problems: they don’t work well across different situations or platforms, and they’re not good at predicting what will happen in real-life situations. This makes it hard to use them in many different industries. The new FOUND model solves these problems by combining data from many different scenarios and aligning it with the demands of each scenario. It also uses information about what people might do in the future to make better predictions. Tests show that FOUND is much better than existing methods at targeting users across different situations and platforms. It’s even been used on a real platform called Alipay, where it has been very useful. |
Keywords
» Artificial intelligence » Transferability
