Predicting clicks: estimating the click-through rate for new ads

By Matthew Richardson, Ewa Dominowska, Robert J. Ragno. The Web Conference, 2007

1,253 citations recorded by the source corpus.

Subfields: Information Systems, Marketing

View Paper: https://www.semanticscholar.org/paper/505253630ab7e8f35e26e27904bd3c8faea3c5ce

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Information Systems

In Marketing

Citations in the corpus

Deep stacking networks for information retrieval (2013, background): “, at the top or right hand side of the web search results [11][9].”
Click Models for Web Search (2015, methodology): “Predicting clicks in sponsored search [Richardson et al., 2007] has an immense influence on revenue and is often done using click models (see, e.”
Position Bias Estimation for Unbiased Learning to Rank in Personal Search (2018, methodology): “The position bias model [27] and the Cascade model [10] are the two classic click models.”
Deep Learning over Multi-field Categorical Data - - A Case Study on User Response Prediction (2016, methodology): “The majority of current models use logistic regression based on a set of sparse binary features converted from the original categorical features via one-hot encoding [26,32].”
Practical Lessons from Predicting Clicks on Ads at Facebook (2014, neutral): “So, are you question?”
Learning to Rank with Selection Bias in Personal Search (2016, methodology): “Noticeably, our selection bias estimation methods are related to the position bias model in [30].”
Online Models for Content Optimization (2008, background): “Recent work on matching ads to queries [10] and ads to webpages [1] ar e related.”
RecBole: Towards a Unified, Comprehensive and Efficient Framework for Recommendation Algorithms (2020, background): “C on te xt -a w ar e re co m m en da tio n LR [53] WWW 2007 Microsoft web search dataset”
Predicting accurate probabilities with a ranking loss (2012, background): “For example, when studying the click behaviour of ads in computational advertising, it is essential to model the probability of an ad being clicked, rather than just predicting whether or not it will be clicked (Richardson et al., 2007).”
Tailoring click models to user goals (2009, background): “[14] suggested the examination hypothesis as a general solution to account for the position bias, under which the chance of a click is decoupled to two factors: the examination probability and the document relevance.”
Optimized Cost per Click in Taobao Display Advertising (2017, methodology): “Logistic Regression is a widely used algorithm in tasks like CTR prediction [19].”
Enlister: baidu's recommender system for the biggest chinese Q&A website (2012, methodology): “As in many other information retrieval [2] and search engine advertising application [3] [4], CTR prediction is introduced to improve the performance of the Q&A RSs [5] [6].”
Practical Privacy Preserving POI Recommendation (2020, background): “, Microsoft [49], Google [36], and Ant Financial [12], due to its simplicity, scalability, and online learning capability.”
Sequential Click Prediction for Sponsored Search with Recurrent Neural Networks (2014, methodology): “…historical CTR for ad impressions 2 with respect to different elements, e.g., CTR of query, ad, user, and their combinations (Graepel et al. 2010; Richardson, Dominowska, and Ragno 2007); 2) semantic relevance between query and ad (Radlinski et al. 2008; Shaparenko, Çetin, and Iyer 2009;…”
Large-Scale Learning with Less RAM via Randomization (2013, background): “…true for training massive-scale distributed learning systems, such as those used for predicting ad click through rates (CTR) for sponsored search (Richardson et al., 2007; Craswell et al., 2008; Bilenko & Richardson, 2011; Streeter & McMahan, 2010) or for ﬁltering email spam at scale (Goodman et…”
Overlapping experiment infrastructure: more, better, faster experimentation (2010, background): “wants to test a new machine learning algorithm for predicting the CTR of ads, or even to test variations of an existing algorithm (e.g, by adjusting the learning or shrinkage rate) [1, 10].”
Simple and Scalable Response Prediction for Display Advertising (2014, methodology): “Predicting user response with sparse data has been the focus of However, most of this work was conducted on search advertising and therefore utilized advertisers bid on [Regelson and Fain 2006; Richardson et al. 2007].”