RecSys: Rajeev Rastogi on three recommendation system challenges

In a keynote address, the Amazon International vice president will discuss recommendations in directed graphs, training models whose target labels change, and using prediction uncertainty to improve model performance.

Rajeev Image 2.jpg
Rajeev Rastogi, vice president of applied science in Amazon’s International Emerging Stores division.

In a keynote address at this year’s ACM Conference on Recommender Systems (RecSys), which starts next week, Rajeev Rastogi, vice president of applied science in Amazon’s International Emerging Stores division, will discuss three problems his organization has faced in its work on recommendation algorithms: recommendations in directed graphs; training machine learning models when target labels change over time; and leveraging estimates of prediction uncertainty to improve models’ accuracy.

“The connections are that these are general techniques that cut across many different recommendation problems,” Rastogi explains. “And these are things that we actually use in practice. They make a difference in the real world.”

Directed graphs

The first problem involves directed graphs, or graphs whose edges describe relationships that run in only one direction.

“Directed graphs have applications in many different domains out there — from citation networks, where an edge U-V indicates paper U cites paper V, or in social networks, where an edge U-V would show that user U follows another user V, and in e-commerce, where an edge U-V indicates that customers bought product U before they bought product V,” Rastogi explains.

Although the problem of exploring directed graphs is general, the researchers in Rastogi’s organization focused on this last case: related-products recommendation, where the goal is to predict what other products might interest a customer who has just made a purchase.

“The interesting part here is that the related-products relationship is actually asymmetric,” Rastogi explains. “If you have, say, two nodes, a phone and a phone case, given a phone, you want to recommend a phone case. But if the customer has bought a phone case, you don't want to recommend a phone, because they most likely already have one.”

Like many graph-based applications, the Amazon team’s solution to the problem of asymmetric related-product recommendation involves graph neural networks (GNNs), in which each node of a graph is embedded in a representational space where geometric relationships between nodes carry information about their relationships in the network. The embedding process is iterative, with each iteration factoring in information about nodes at greater removes, until each node’s embedding carries information about its neighborhood.

“A single embedding space does not have the expressive power to model the asymmetric relationships between nodes in directed graphs,” Rastogi explains. “Something that we borrowed from past work is to represent each node with dual embeddings, and one of our novel contributions is really to learn these dual embeddings in a GNN setting that leverages the entire graph structure.”

BLADE.png
At center is a graph indicating the relationships between cell phones and related products such as a case, a power adaptor, and a screen guard. At left is a schematic illustrating the embedding (vector representation) of node A in a traditional graph neural network (GNN); at right is the dual embedding of A, as both a recommendation target (A-t) and a recommendation source (A-s), in BLADE. From "BLADE: Biased neighborhood sampling based graph neural network for directed graphs".

“Then we had additional techniques, like adaptive sampling,” Rastogi adds. “These vanilla GNNs sample fixed neighborhood sizes for every node. But we found that low-degree nodes” — that is, nodes with few connections to other nodes — “have suboptimal performance when you have fixed neighborhood sizes for every node, because low-degree nodes have sparse connectivity structures. And so less information gets transmitted when you're aggregating information from neighbors and so on.

“So we actually choose to sample larger neighborhoods for low-degree nodes and smaller neighborhoods for high-degree nodes. It's a little bit counterintuitive, but it gives us much better results.”

Delayed feedback

A typical machine learning (ML) model is trained on labeled data, and the model must learn to predict the labels — its training targets — from the data. The second problem Rastogi addresses in his talk is how best to train a model when you know that some of the target labels are going to change in the near future.

“This is, again, a very common problem across many different domains,” Rastogi says. “In recommendations, there can be a time lag of a few days between customers viewing a recommendation and purchasing the product.

“There's a trade-off here: If you use all the training data in real time, some of those more recent training examples may have target labels that are incorrect, because they are going to change over time. On the other hand, if you ignore all the training examples you got in the last five days, then you're missing out on recent data, and your model isn't going to be as good — especially in environments where models need to be retrained frequently.

Delayed feedback.png
An illustration of true negatives, delayed positives and true positives, from "Modelling delayed redemption with importance sampling and pre-redemption engagement".

“Here, what we've done is come up with an importance-sampling strategy that essentially weighs every training example with an importance weight. Let P(X,Y) be the true data distribution, and Q(X,Y) be the data distribution that you observe in the training set. Our importance-sampling strategy uses the ratio P(X,Y) divided by Q(X,Y) as the importance weight.

“Our key innovation centers on techniques to compute these importance weights in new scenarios. One is where we take into account preconversion signals. People tend to do something before they convert; they may add to cart, or they may click on the product to research it before completing the purchase. So we take into account those signals, and that helps us overcome data sparsity.

“But then it makes the computation of importance weights a little bit more complex. If it's very likely that the target label will actually change from 0 — a negative example — to 1 , then the importance weight would be much lower than if the likelihood of the example not changing was very low. Essentially, what you're trying to do is learn from the data the likelihood that the target label is going is change in the future and capture that in the importance weights.”

Prediction uncertainty

Finally, Rastogi says, the third technique he’ll discuss in his talk is the use of uncertainty estimates to improve the accuracy of model predictions.

“ML models typically will return point estimates,” Rastogi explains. “But usually you have a probability distribution. In some cases, you could know there's a 0.5 chance this customer is going buy the product. But in some cases, it could be anywhere between 0.2 and 0.8. What we found is, if you’re able to generate uncertainty estimates for model predictions, we can exploit them to improve model accuracy.

“We trained a binary classifier to predict ad click probability for an ads recommendation application. For every sample in the holdout set, we generated both the model score, which is the probability prediction, and also an uncertainty estimate, which is how certain I am about the predicted probability.

“If I looked at a lot of examples in the holdout set with a model score of 0.5, you would expect that about 50% of them resulted in clicks: that’s the empirical positivity rate. If it were 0.8, then the empirical positivity rate should be around 80%.

“But what we found is that as the variance of the model score increased, the empirical positivity rates went down. If I have a score of 0.8, I could say, well, it's between 0.79 and 0.81, which corresponds to a low variance. Or I could say, it's between 0.65 and 0.95, which indicates a high variance. We found that for the same model score, as the confidence intervals became larger, the empirical positivity rate started dropping.

“That has implications on selecting the decision boundary for binary classifiers. Traditionally, binary classifiers used a single threshold on model scores. But now, since the empirical positivity rate depends on both the model score and the uncertainty estimate, just selecting a single threshold value turns out to be suboptimal. If we select multiple thresholds, one per uncertainty level, we found that we can get much higher recall for a given precision.”

Members of Rastogi’s organization are currently writing a paper on their prediction uncertainty work — but the method is already in production.

“There are a lot of things that people publish papers about, and they're forgotten and never really used,” Rastogi says. “Coming from Amazon, we do science that actually makes a difference to customers and solves customer pain points. These are three examples of doing customer-obsessed science that actually makes a difference in the real world.”

Related content

US, CA, Santa Clara
Machine learning (ML) has been strategic to Amazon from the early years. We are pioneers in areas such as recommendation engines, product search, eCommerce fraud detection, and large-scale optimization of fulfillment center operations. The Generative AI team helps AWS customers accelerate the use of Generative AI to solve business and operational challenges and promote innovation in their organization. As an applied scientist, you are proficient in designing and developing advanced ML models to solve diverse challenges and opportunities. You will be working with terabytes of text, images, and other types of data to solve real-world problems. You'll design and run experiments, research new algorithms, and find new ways of optimizing risk, profitability, and customer experience. We’re looking for talented scientists capable of applying ML algorithms and cutting-edge deep learning (DL) and reinforcement learning approaches to areas such as drug discovery, customer segmentation, fraud prevention, capacity planning, predictive maintenance, pricing optimization, call center analytics, player pose estimation, event detection, and virtual assistant among others. AWS Sales, Marketing, and Global Services (SMGS) is responsible for driving revenue, adoption, and growth from the largest and fastest growing small- and mid-market accounts to enterprise-level customers including public sector. The AWS Global Support team interacts with leading companies and believes that world-class support is critical to customer success. AWS Support also partners with a global list of customers that are building mission-critical applications on top of AWS services. Key job responsibilities The primary responsibilities of this role are to: Design, develop, and evaluate innovative ML models to solve diverse challenges and opportunities across industries Interact with customer directly to understand their business problems, and help them with defining and implementing scalable Generative AI solutions to solve them Work closely with account teams, research scientist teams, and product engineering teams to drive model implementations and new solutions About the team Diverse Experiences AWS values diverse experiences. Even if you do not meet all of the qualifications and skills listed in the job description, we encourage candidates to apply. If your career is just starting, hasn’t followed a traditional path, or includes alternative experiences, don’t let it stop you from applying. Why AWS? Amazon Web Services (AWS) is the world’s most comprehensive and broadly adopted cloud platform. We pioneered cloud computing and never stopped innovating — that’s why customers from the most successful startups to Global 500 companies trust our robust suite of products and services to power their businesses. Inclusive Team Culture Here at AWS, it’s in our nature to learn and be curious. Our employee-led affinity groups foster a culture of inclusion that empower us to be proud of our differences. Ongoing events and learning experiences, including our Conversations on Race and Ethnicity (CORE) and AmazeCon (gender diversity) conferences, inspire us to never stop embracing our uniqueness. Mentorship & Career Growth We’re continuously raising our performance bar as we strive to become Earth’s Best Employer. That’s why you’ll find endless knowledge-sharing, mentorship and other career-advancing resources here to help you develop into a better-rounded professional. Work/Life Balance We value work-life harmony. Achieving success at work should never come at the expense of sacrifices at home, which is why flexible work hours and arrangements are part of our culture. When we feel supported in the workplace and at home, there’s nothing we can’t achieve in the cloud. We are open to hiring candidates to work out of one of the following locations: San Francisco, CA, USA | Santa Clara, CA, USA
US, WA, Bellevue
We are seeking a passionate, talented, and inventive individual to join the Amazon AGI team and help build industry-leading technologies that customers will love. This team offers a unique opportunity to make a significant impact on the customer experience and contribute to the design, architecture, and implementation of a cutting-edge product. Our goal is to provide conversational, visual and proactive mechanisms to delight customers, increase customer engagement, reduce defects, and enable natural interactions across Amazon touch points (e.g., Alexa, Stores, Mobile). To achieve our mission, we build automated, scalable, self-serve AI systems that use customer, device and ambient signals to offer personalized suggestions, comprehend customer inputs, learn from customer interactions, reduce defects, and propose appropriate actions to serve millions of Amazon customers across the globe. Key job responsibilities As a Principal Applied Scientist, you will be a trusted part of the technical leadership. You bring business and industry context to science and technology decisions. You tackle intrinsically hard problems, acquiring expertise as needed and decompose complex problems into straightforward solutions. You align teams toward coherent strategies. You will develop long-term strategies, persuade teams to adopt those strategies, propose goals and deliver on them. You will also participate in organizational planning, hiring, mentorship and leadership development. You will serve as a key scientific resource in full-cycle development (conception, design, implementation, testing to documentation, delivery, and maintenance). About the team The Artificial General Intelligence (AGI) team is looking for a passionate, talented, and inventive Principal Scientist to help build industry-leading conversational technologies that customers love. Our mission is to push the envelope in Artificial Intelligence (AI), Natural Language Understanding (NLU), Machine Learning (ML), Dialog Management, Automatic Speech Recognition (ASR), and Audio Signal Processing, in order to provide the best-possible experience for our customers We are open to hiring candidates to work out of one of the following locations: Bellevue, WA, USA | Boston, MA, USA | Seattle, WA, USA | Sunnyvale, CA, USA
US, NY, New York
We are looking for a motivated and experienced Senior Data Scientist with experience in Machine Learning (ML), Artificial Intelligence (AI), Big Data, and Service Oriented Architecture with deep understanding in advertising businesses, to be part of a team of talented scientists and engineers to innovate, iterate, and solve real world problem with cutting-edge AWS technologies. In this role, you will take a leading role in defining the problem, innovating the ML/AI solutions, and information the tech roadmap. You will join a cross-functional, fun-loving team, working closely with scientists and engineers in a daily basis. You will innovate on behalf of our customers by prototyping, delivering functional proofs of concept (POCs), and partnering with our engineers to productize and scale successful POCs. If you are passionate about creating the future, come join us as we have fun, and make history. Key job responsibilities - Define and execute a research & development roadmap that drives data-informed decision making for marketers and advertisers - Establish and drive data hygiene best practices to ensure coherence and integrity of data feeding into production ML/AI solutions - Collaborate with colleagues across science and engineering disciplines for fast turnaround proof-of-concept prototyping at scale - Partner with product managers and stakeholders to define forward-looking product visions and prospective business use cases - Drive and lead of culture of data-driven innovations within and outside across Amazon Ads Marketing orgs About the team Marketing Decision Science provides science products to enable Amazon Ads Marketing to deliver relevant and compelling guidance across marketing channels to prospective and active advertisers for success on Amazon. We own the product, technology and deployment roadmap for AI- and analytics-powered products across Amazon Ads Marketing. We analyze the needs, experiences, and behaviors of Amazon advertisers at petabytes scale, to deliver the right marketing communications to the right advertiser at the right team to help them make the data-informed advertising decisions. Our science-based products enable applications and synergies across Ads organization, spanning marketing, product, and sales use cases. We are open to hiring candidates to work out of one of the following locations: New York, NY, USA
US, WA, Bellevue
We are looking for detail-oriented, organized, and responsible individuals who are eager to learn how to work with large and complicated data sets. Some knowledge of econometrics, as well as basic familiarity with Python is necessary, and experience with SQL and UNIX would be a plus. These are full-time positions at 40 hours per week, with compensation being awarded on an hourly basis. You will learn how to build data sets and perform applied econometric analysis at Internet speed collaborating with economists, scientists, and product managers. These skills will translate well into writing applied chapters in your dissertation and provide you with work experience that may help you with placement. Roughly 85% of previous cohorts have converted to full time scientist employment at Amazon. If you are interested, please send your CV to our mailing list at econ-internship@amazon.com. We are open to hiring candidates to work out of one of the following locations: Bellevue, WA, USA
US, WA, Seattle
Are you excited about developing models to revolutionize automation, robotics and computer vision? Are you looking for opportunities to build and deploy them on real problems at truly vast scale? At Amazon Fulfillment Technologies and Robotics we are on a mission to build high-performance autonomous systems that perceive and act to further improve our world-class customer experience - at Amazon scale. We are looking for scientists, engineers and program managers for a variety of roles. The Amazon Robotics software team is seeking a collaborative Applied Scientist to focus on computer vision machine learning models. This includes building multi-viewpoint and time-series computer vision systems. It includes building large-scale models using data from many different tasks and scenes. This work spans from basic research such as cross domain training, to experimenting on prototype in the lab, to running wide-scale A/B tests on robots in our facilities. Key job responsibilities * Research vision - Where should we be focusing our efforts * Research delivery – Proving/dis-proving strategies in offline data or in the lab * Production studies - Insights from production data or ad-hoc experimentation. A day in the life Amazon offers a full range of benefits that support you and eligible family members, including domestic partners and their children. Benefits can vary by location, the number of regularly scheduled hours you work, length of employment, and job status such as seasonal or temporary employment. The benefits that generally apply to regular, full-time employees include: 1. Medical, Dental, and Vision Coverage 2. Maternity and Parental Leave Options 3. Paid Time Off (PTO) 4. 401(k) Plan If you are not sure that every qualification on the list above describes you exactly, we'd still love to hear from you! At Amazon, we value people with unique backgrounds, experiences, and skillsets. If you’re passionate about this role and want to make an impact on a global scale, please apply! We are open to hiring candidates to work out of one of the following locations: Seattle, WA, USA
US, WA, Bellevue
Looking for your next challenge? North America Sort Centers (NASC) are experiencing growth and looking for a skilled, highly motivated Data Scientist to join the NASC Engineering Data, Product and Simulation Team. The Sort Center network is the critical Middle-Mile solution in the Amazon Transportation Services (ATS) group, linking Fulfillment Centers to the Last Mile. The experience of our customers is dependent on our ability to efficiently execute volume flow through the middle-mile network. Key job responsibilities The Senior Data Scientist will design and implement solutions to address complex business questions using simulation. In this role, you will apply advanced analysis techniques and statistical concepts to draw insights from massive datasets, and create intuitive simulations and data visualizations. You can contribute to each layer of a data solution – you work closely with process design engineers, business intelligence engineers and technical product managers to obtain relevant datasets and create simulation models, and review key results with business leaders and stakeholders. Your work exhibits a balance between scientific validity and business practicality. On this team, you will have a large impact on the entire NASC organization, with lots of opportunity to learn and grow within the NASC Engineering team. This role will be the first dedicated simulation expert, so you will have an exceptional opportunity to define and drive vision for simulation best practices on our team. To be successful in this role, you must be able to turn ambiguous business questions into clearly defined problems, develop quantifiable metrics and deliver results that meet high standards of data quality, security, and privacy. About the team NASC Engineering’s Product and Analytics Team’s sole objective is to develop tools for under the roof simulation and optimization, supporting the needs of our internal and external stakeholders (i.e Process Design Engineering, NASC Engineering, ACES, Finance, Safety and Operations). We develop data science tools to evaluate what-if design and operations scenarios for new and existing sort centers to understand their robustness, stability, scalability, and cost-effectiveness. We conceptualize new data science solutions, using optimization and machine learning platforms, to analyze new and existing process, identify and reduce non-value added steps, and increase overall performance and rate. We work by interfacing with various functional teams to test and pilot new hardware/software solutions. We are open to hiring candidates to work out of one of the following locations: Bellevue, WA, USA
IN, KA, Bangalore
Alexa is the voice activated digital assistant powering devices like Amazon Echo, Echo Dot, Echo Show, and Fire TV, which are at the forefront of this latest technology wave. To preserve our customers’ experience and trust, the Alexa Sensitive Content Intelligence (ASCI) team creates policies and builds services and tools through Machine Learning techniques to detect and mitigate sensitive content across Alexa. We are looking for an experienced Senior Applied Scientist to build industry-leading technologies in attribute extraction and sensitive content detection across all languages and countries. An Applied Scientist will be a tech lead for a team of exceptional scientists to develop novel algorithms and modeling techniques to advance the state of the art in NLP or CV related tasks. You will work in a hybrid, fast-paced organization where scientists, engineers, and product managers work together to build customer facing experiences. You will collaborate with and mentor other scientists to raise the bar of scientific research in Amazon. Your work will directly impact our customers in the form of products and services that make use of speech, language, and computer vision technologies. We are looking for a leader with strong technical experiences a passion for building scientific driven solutions in a fast-paced environment. You should have good understanding of NLP models (e.g. LSTM, transformer based models) or CV models (e.g. CNN, AlexNet, ResNet) and where to apply them in different business cases. You leverage your exceptional technical expertise, a sound understanding of the fundamentals of Computer Science, and practical experience of building large-scale distributed systems to creating reliable, scalable, and high-performance products. In addition to technical depth, you must possess exceptional communication skills and understand how to influence key stakeholders. You will be joining a select group of people making history producing one of the most highly rated products in Amazon's history, so if you are looking for a challenging and innovative role where you can solve important problems while growing as a leader, this may be the place for you. Key job responsibilities You'll lead the science solution design, run experiments, research new algorithms, and find new ways of optimizing customer experience. You set examples for the team on good science practice and standards. Besides theoretical analysis and innovation, you will work closely with talented engineers and ML scientists to put your algorithms and models into practice. Your work will directly impact the trust customers place in Alexa, globally. You contribute directly to our growth by hiring smart and motivated Scientists to establish teams that can deliver swiftly and predictably, adjusting in an agile fashion to deliver what our customers need. A day in the life You will be working with a group of talented scientists on researching algorithm and running experiments to test scientific proposal/solutions to improve our sensitive contents detection and mitigation. This will involve collaboration with partner teams including engineering, PMs, data annotators, and other scientists to discuss data quality, policy, and model development. You will mentor other scientists, review and guide their work, help develop roadmaps for the team. You work closely with partner teams across Alexa to deliver platform features that require cross-team leadership. About the hiring group About the team The mission of the Alexa Sensitive Content Intelligence (ASCI) team is to (1) minimize negative surprises to customers caused by sensitive content, (2) detect and prevent potential brand-damaging interactions, and (3) build customer trust through appropriate interactions on sensitive topics. The term “sensitive content” includes within its scope a wide range of categories of content such as offensive content (e.g., hate speech, racist speech), profanity, content that is suitable only for certain age groups, politically polarizing content, and religiously polarizing content. The term “content” refers to any material that is exposed to customers by Alexa (including both 1P and 3P experiences) and includes text, speech, audio, and video. We are open to hiring candidates to work out of one of the following locations: Bangalore, KA, IND
US, WA, Seattle
Amazon Web Services (AWS) is building a world-class marketing organization, and we are looking for an experienced Applied Scientist to join the central data and science organization for AWS Marketing. You will lead AWS Measurement, targeting, recommendation, forecasting related AI/ML products and initiatives, and own mechanisms to raise the science and measurement standard. You will work with economists, scientists and engineers within the team, and partner with product and business teams across AWS Marketing to build the next generation marketing measurement, valuation and machine learning capabilities directly leading to improvements in our key performance metrics. A successful candidate has an entrepreneurial spirit and wants to make a big impact on AWS growth. You will develop strong working relationships and thrive in a collaborative team environment. You will work closely with business leaders, scientists, and engineers to translate business and functional requirements into concrete deliverables, including the design, development, testing, and deployment of highly scalable distributed services. The ideal candidate will have experience with machine learning models and causal inference. Additionally, we are seeking candidates with strong rigor in applied sciences and engineering, creativity, curiosity, and great judgment. You will work on high-impact, high-visibility products, with your work improving the experience of AWS leads and customers. Key job responsibilities - Lead the design, development, deployment, and innovation of advanced science models in the strategic area of marketing measurement and optimization. - Partner with scientists, economists, engineers, and product leaders to break down complex business problems into science approaches. - Understand and mine the large amount of data, prototype and implement new learning algorithms and prediction techniques to improve long-term causal estimation approaches. - Design, build, and deploy effective and innovative ML solutions to improve components of our ML and causal inference pipelines. - Publish and present your work at internal and external scientific venues in the fields of ML and causal inference. - Influence long-term science initiatives and mentor other scientists across AWS. We are open to hiring candidates to work out of one of the following locations: Arlington, VA, USA | Austin, TX, USA | New York, NY, USA | Santa Clara, CA, USA | Seattle, WA, USA
US, WA, Bellevue
Where will Amazon's growth come from in the next year? What about over the next five? Which product lines are poised to quintuple in size? Are we investing enough in our infrastructure, or too much? How do our customers react to changes in prices, product selection, or delivery times? These are among the most important questions at Amazon today. The Topline Forecasting team in the Supply Chain Optimization Technologies (SCOT) group is looking for innovative, passionate and results-oriented Principal Economist to provide thought-leadership to help answer these questions. You will have an opportunity to own the long-run outlook for Amazon’s global consumer business and shape strategic decisions at the highest level. The successful candidate will be able to formalize problem definitions from ambiguous requirements, build econometric models using Amazon’s world-class data systems, and develop cutting-edge solutions for non-standard problems. Key job responsibilities • You understand the state-of-the-art in time series and econometric modeling. • You apply econometric tools and theory to solve business problems in a fast moving environment. • You excel at extracting insights and correct interpretations from data using advanced modeling techniques. • You communicate insights in a digestible manner to senior leaders in Finance and Operations within the company. • You are able to anticipate future business challenges and key questions, and have the ability to design modeling solutions to tackle them. • You have broad influence over the Topline team’s scientific research agenda and deliverables. • You contribute to the broader Econ research community in Amazon. • You advise other economists on scientific best-practices and raise the bar of research. • You will actively mentor other scientists and contribute to their career development. We are open to hiring candidates to work out of one of the following locations: Bellevue, WA, USA | New York, NY, USA
US, WA, Seattle
Innovators wanted! Are you an entrepreneur? A builder? A dreamer? This role is part of an Amazon Special Projects team that takes the company’s Think Big leadership principle to the extreme. We focus on creating entirely new products and services with a goal of positively impacting the lives of our customers. No industries or subject areas are out of bounds. If you’re interested in innovating at scale to address big challenges in the world, this is the team for you. Here at Amazon, we embrace our differences. We are committed to furthering our culture of inclusion. We have thirteen employee-led affinity groups, reaching 40,000 employees in over 190 chapters globally. We are constantly learning through programs that are local, regional, and global. Amazon’s culture of inclusion is reinforced within our 16 Leadership Principles, which remind team members to seek diverse perspectives, learn and be curious, and earn trust. Our team highly values work-life balance, mentorship and career growth. We believe striking the right balance between your personal and professional life is critical to life-long happiness and fulfillment. We care about your career growth and strive to assign projects and offer training that will challenge you to become your best. We are open to hiring candidates to work out of one of the following locations: Seattle, WA, USA