Alex Shtoff

Alex Shtoff

Auto-generated profile, based on publicly available information

Alex Shtoff

Yahoo Research, Haifa, Israel

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
13 May 2024
Verified
Conference Paper
Open access
RSRotem Stram
Rotem Stram
RARani Abboud
Rani Abboud
ASAlex Shtoff
Alex Shtoff
OSOren Somekh
Oren Somekh
ARAriel Raviv
Ariel Raviv
YKYair Koren
Yair Koren

8 October 2024
Verified
Conference Paper
Open access
ASAlex Shtoff
Alex Shtoff
MVMichael Viderman
Michael Viderman
NHNaama Haramaty-Krasne
Naama Haramaty-Krasne
OSOren Somekh
Oren Somekh
ARAriel Raviv
Ariel Raviv
TBTularam Ban
Tularam Ban
Factorization machine (FM) variants are widely used in recommendation systems that operate under strict throughput and latency requirements, such as online advertising systems. FMs have two prominent strengths. First, is their ability to model pairwise feature interactions while being resilient to data sparsity by learning factorized representations. Second, their computational graphs facilitate fast inference and training. Moreover, when items are ranked as a part of a query for each incoming user, these graphs facilitate computing the portion stemming from the user and context fields only once per query. Thus, the computational cost for each ranked item is proportional only to the number of fields that vary among the ranked items. Consequently, in terms of inference cost, the number of user or context fields is practically unlimited. More advanced variants of FMs, such as field-aware and field-weighted FMs, provide better accuracy by learning a representation of field-wise interactions, but require computing all pairwise interaction terms explicitly. In particular, the computational cost during inference is proportional to the square of the number of fields, including user, context, and item. When the number of fields is large, this is prohibitive in systems with strict latency constraints, and imposes a limit on the number of user and context fields for a given computational budget. To mitigate this caveat, heuristic pruning of low intensity field interactions is commonly used to accelerate inference. In this work we propose an alternative to the pruning heuristic in field-weighted FMs using a diagonal plus symmetric low-rank decomposition. Our technique reduces the computational cost of inference, by allowing it to be proportional to the number of item fields only. Using a set of experiments on real-world datasets, we show that aggressive rank reduction outperforms similarly aggressive pruning in both accuracy and item recommendation speed. Beyond computational complexity analysis, we corroborate our claim of faster inference experimentally, both via a synthetic test, and by having deployed our solution to a major online advertising system, where we observed significant ranking latency improvements. We have made the code to reproduce the results on public datasets and synthetic tests available at https://github.com/michaelviderman/pytorch-fm.

1 August 2024 in Mathematical Programming Computation
Verified
Journal Article
Closed access
PDF file
ASAlex Shtoff
Alex Shtoff

19 July 2026
Verified
Conference Paper
Open access
DCDavid Carmel
David Carmel
SFSimone Filice
Simone Filice
GHGuy Horowitz
Guy Horowitz
YMYoelle Maarek
Yoelle Maarek
ASAlexander Shtoff
Alexander Shtoff
RTRan Tavory
Ran Tavory

13 November 2022 in arXiv
Verified
Preprint
Open access
PDF file
YKYohay Kaplan
Yohay Kaplan
YKYair Koren
Yair Koren
ASAlex Shtoff
Alex Shtoff
TSTomer Shadi
Tomer Shadi
OSOren Somekh
Oren Somekh
Yahoo Gemini native advertising marketplace serves billions of impressions daily, to hundreds millions of unique users, and reaches a yearly revenue of many hundreds of millions USDs. Powering Gemini native models for predicting advertise (ad) event probabilities, such as conversions and clicks, is OFFSET - a feature enhanced collaborative-filtering (CF) based event prediction algorithm. The predicted probabilities are then used in Gemini native auctions to determine which ads to present for every serving event (impression). Dynamic creative optimization (DCO) is a recent Gemini native product that was launched two years ago and is increasingly gaining more attention from advertisers. The DCO product enables advertisers to issue several assets per each native ad attribute, creating multiple combinations for each DCO ad. Since different combinations may appeal to different crowds, it may be beneficial to present certain combinations more frequently than others to maximize revenue while keeping advertisers and users satisfied. The initial DCO offer was to optimize click-through rates (CTR), however as the marketplace shifts more towards conversion based campaigns, advertisers also ask for a {conversion based solution. To accommodate this request, we present a post-auction solution, where DCO ads combinations are favored according to their predicted conversion rate (CVR). The predictions are provided by an auxiliary OFFSET based combination CVR prediction model, and used to generate the combination distributions for DCO ad rendering during serving time. An online evaluation of this explore-exploit solution, via online bucket A/B testing, serving Gemini native DCO traffic, showed a 53.5% CVR lift, when compared to a control bucket serving all combinations uniformly at random.

25 January 2024 in arXiv
Verified
Preprint
Open access
PDF file
ASAlex Shtoff
Alex Shtoff
YKYohay Kaplan
Yohay Kaplan
ARAriel Raviv
Ariel Raviv
The task of predicting conversion rates (CVR) lies at the heart of online advertising systems aiming to optimize bids to meet advertiser performance requirements. Even with the recent rise of deep neural networks, these predictions are often made by factorization machines (FM), especially in commercial settings where inference latency is key. These models are trained using the logistic regression framework on labeled tabular data formed from past user activity that is relevant to the task at hand. Many advertisers only care about click-attributed conversions. A major challenge in training models that predict conversions-given-clicks comes from data sparsity - clicks are rare, conversions attributed to clicks are even rarer. However, mitigating sparsity by adding conversions that are not click-attributed to the training set impairs model calibration. Since calibration is critical to achieving advertiser goals, this is infeasible. In this work we use the well-known idea of self-supervised pre-training, and use an auxiliary auto-encoder model trained on all conversion events, both click-attributed and not, as a feature extractor to enrich the main CVR prediction model. Since the main model does not train on non click-attributed conversions, this does not impair calibration. We adapt the basic self-supervised pre-training idea to our online advertising setup by using a loss function designed for tabular data, facilitating continual learning by ensuring auto-encoder stability, and incorporating a neural network into a large-scale real-time ad auction that ranks tens of thousands of ads, under strict latency constraints, and without incurring a major engineering cost. We show improvements both offline, during training, and in an online A/B test. Following its success in A/B tests, our solution is now fully deployed to the Yahoo native advertising system.