On benchmark data set, question-answering system halves error rate

Improvements come from new transfer learning method, new publicly released data set.

The last few years have seen great advances in the design of language models, which are a critical component of language-based AI systems. Language models can be used to compute the probability of any given sequence (even discontinuous sequences) of words, which is useful in natural-language processing.

The new language models are all built atop the Transformer neural architecture, which is particularly good at learning long-range dependencies among input data, such as the semantic and syntactic relationships between individual words of a sentence. At the annual meeting of the Association for the Advancement of Artificial Intelligence, my colleagues and I will present a method for adapting these new models — BERT, for instance — to the problem of answer selection, a central topic in the field of question answering.

In tests on an industry-standard benchmark data set, our new model demonstrated a 10% absolute improvement in mean average precision over the previous state-of-the-art answer selection model. That translates to an error rate reduction of 50%.

Our approach uses transfer learning, in which a machine learning model pretrained on one task — here, word sequence prediction — is fine-tuned on another — here, answer selection. Our innovation is to introduce an intermediate step between the pre-training of the source model and its adaptation to new target domains.

In the intermediate step, we fine-tune the language model on a large corpus of general question-answer pairs. Then we fine-tune it even further, on a small body of topic-specific questions and answers (the target domain). We call our system TANDA, for transfer and adapt.

The corpus we use for the intermediate step is based on the public data set Natural Questions (NQ), which was designed for the training of reading comprehension systems. We transformed NQ so that it could be used to train answer selection systems instead, and the public release of our modified data set — dubbed ASNQ, for answer selection NQ — is itself an important contribution to the research community.

BERT transfer and adapt
The transfer and adapt stages of TANDA, which give it its name.

In addition to its performance gains, our system has several other advantages:

1. It can be fine-tuned on target data without a heavy hyperparameter search. Hyperparameters are characteristics of a neural network such as the number of layers, the number of nodes per layer, and the learning rate of the training algorithm, which are often determined through trial and error. The stability of our model means that it can be adapted to a target domain with very little training data.

2. It’s robust to noise: in our tests, errors in the target domain data had little effect on the system’s accuracy. Again, this is important because of the difficulty of acquiring high-quality data.

3. The most time-consuming part of our procedure — the intermediate step — need be performed only once. The resulting model can be adapted to an indefinite number of target domains.

Nuts and bolts

Answer selection assumes that for a given question, the system has access to a set of candidate answers; in practice, candidates are often assembled through standard keyword search. Answer selection systems are thus trained on pairs of sentences — one question and one candidate answer at a time — and try to learn which candidates are viable answers.

In the past, researchers have attempted to adapt Transformer-based language models to answer selection by directly fine-tuning them on small sets of domain-specific data, but we hypothesized that the addition of an intermediate fine-tuning step would lead to better results.

In reading comprehension, the system receives a question and a block of text; its job is to select the one sentence in the block that best answers the question. NQ is a set of text blocks in each of which one sentence has been labeled as the best answer. To convert NQ into the answer selection data set ASNQ, we extracted the best answers from their text blocks and labeled them as successful answer sentences. The other sentences in each block we labeled unsuccessful answers.

Generally, Transformer-based language models are trained on sentences in which particular words have been hidden, or “masked”, and the models must learn to fill in the blanks.

BERT pretraining
An example of a BERT training input, consisting of two masked sentences, A and B. Each sentence consists of a series of “tokens” (Tok 1 – Tok N and Tok 1 – Tok M), a term that takes in both words and other symbols. The sentences are preceded by a special start token ([CLS]) and separated by a special separator token ([SEP]).
Stacy Reilly

BERT models are also trained on a second objective, which is to determine whether a second input sentence follows naturally from the first. The input to a BERT model is thus a pair of (masked) sentences.

This suits BERT particularly well to the task of answer sentence selection, where the inputs are also pairs of sentences. Our procedure is first to fine-tune a Transformer-based model on ASNQ and then fine-tune it again on a smaller, domain-specific data set — a sports news set, for instance, which might include questions such as “When did the Philadelphia Eagles play the fog bowl?”

We tested our approach using two public data sets, WikiQA and TREC-QA, and evaluated the system’s performance according to mean average precision (MAP) and mean reciprocal rank (MRR). Intuitively, MAP measures the quality of a sorted list of answers according to the correctness of the complete ranking, while MRR measures the probability that the correct answer is near the top of the list.

On WikiQA and TREC-QA, our system’s MAP was 92% and 94.3%, respectively, a significant improvement over the previous records of 83.4% and 87.5%. MRR for our system was 93.3% and 97.4%, up from 84.8% and 94%, respectively.

Related content

US, CA, Santa Clara
Job summaryAmazon is looking for a passionate, talented, and inventive Applied Scientist with a strong machine learning background to help build industry-leading language technology.Our mission is to provide a delightful experience to Amazon’s customers by pushing the envelope in Natural Language Processing (NLP), Natural Language Understanding (NLU), Dialog management, conversational AI and Machine Learning (ML).As part of our AI team in Amazon AWS, you will work alongside internationally recognized experts to develop novel algorithms and modeling techniques to advance the state-of-the-art in human language technology. Your work will directly impact millions of our customers in the form of products and services, as well as contributing to the wider research community. You will gain hands on experience with Amazon’s heterogeneous text and structured data sources, and large-scale computing resources to accelerate advances in language understanding.We are hiring primarily in Conversational AI / Dialog System Development areas: NLP, NLU, Dialog Management, NLG.This role can be based in NYC, Seattle or Palo Alto.Inclusive Team CultureHere at AWS, we embrace our differences. We are committed to furthering our culture of inclusion. We have ten employee-led affinity groups, reaching 40,000 employees in over 190 chapters globally. We have innovative benefit offerings, and host annual and ongoing learning experiences, including our Conversations on Race and Ethnicity (CORE) and AmazeCon (gender diversity) conferences.Work/Life BalanceOur team puts a high value on work-life balance. It isn’t about how many hours you spend at home or at work; it’s about the flow you establish that brings energy to both parts of your life. We believe striking the right balance between your personal and professional life is critical to life-long happiness and fulfillment. We offer flexibility in working hours and encourage you to find your own balance between your work and personal lives.Mentorship & Career GrowthOur team is dedicated to supporting new members. We have a broad mix of experience levels and tenures, and we’re building an environment that celebrates knowledge sharing and mentorship. Our senior members enjoy one-on-one mentoring and thorough, but kind, code reviews. We care about your career growth and strive to assign projects based on what will help each team member develop into a better-rounded engineer and enable them to take on more complex tasks in the future.
IN, TS, Hyderabad
Job summaryAre you excited about driving business growth for millions of sellers by applying Machine Learning? Do you thrive in a fast-moving, large-scale environment that values data-driven decision making and sound scientific practices? We are looking for experienced data scientists to build sophisticated decision making systems that help Amazon Marketplace Sellers to grow their businesses.Amazon Marketplace enables sellers to reach hundreds of millions of customers and provides sellers the tools and services needed to make e-commerce simple, efficient and successful. Our team builds the core intelligence, insights, and algorithms that power a range of products used by millions of sellers. We are tackling large-scale, challenging problems such as helping sellers to prioritise business tasks by bringing together petabytes of data from sources across Amazon.You will be proficient with creating value out of data by formulating questions, analysing vast amounts of data, and communicating insights effectively to audience of varied backgrounds. In addition, you'll contribute to online experiments, build machine learning pipelines and personalised data products.To know more about Amazon science, Please visit https://www.amazon.scienceKey job responsibilities· Collaborate with domain experts, formulate questions, gather, process and analyse petabytes of data to unearth reliable insights· Design & execute experiments and analyze experimental results· Communicate insights effectively to audience of a wide range of backgrounds· Formulate relevant prediction problems and solve them by developing machine learning models· Partner with data engineering teams to improve quality of data assets, metrics and insights· Leverage industry best practices to establish repeatable science practices, principles & processes
US, WA, Seattle
Job summaryAmazon Sub-Same-Day Supply Chain team is looking for an experienced and motivated Senior Data Scientist to generate data-driven insights influencing the long term SSD supply chain strategy, build the necessary predictive models, optimization algorithms and customer behavioral segments allowing us to discover and build the roadmap for SSD to enable operational efficiency and scale.Key job responsibilitiesWork with product managers, engineers, other scientists, and leadership to identify and prioritize complex problems.Translate business problems into specific analytical questions and form hypotheses that can be answered with available data using scientific methods or identify additional data needed in the master datasets to fill any gapsDesign, develop, and evaluate highly innovative statistics and ML modelsGuide and establish scalable, efficient, automated processes for large scale data analyses, model development, model validation and model implementationProactively seek to identify business opportunities and insights and provide solutions to shape key business processes and policies based on a broad and deep knowledge of Amazon data, industry best-practices, and work done by other teams.A day in the lifeIn this role, you will be a technical expert with significant scope and impact. You will work with Product Managers, Business Engineers, and other Scientists, to deeply understand SSDs current optimization strategy while benchmarking against industry best practices and standards to gain insights that will drive our roadmap. A successful Data Scientist will have extreme bias for action needed in a startup environment, with outstanding leadership skills, proven ability to build and manage medium-scale modeling projects, identify data requirements, build methodology and tools that are statistically grounded. It will be a person who enjoys diving deep into data, doing analysis, discovering root causes, and designing long-term scientific solutions. We are seeking someone who can thrive in a fast-paced, high-energy and fun work environment where we deliver value incrementally and frequently. We value highly technical people who know their subject matter deeply and are willing to learn new areas. We look for individuals who know how to deliver results and show a desire to develop themselves, their colleagues, and their career.About the teamAmazon's Sub-Same Day (SSD) delivery program is designed to get customers their items as fast as possible – currently in as quickly as five hours. With ultra-fast delivery becoming increasingly important, we are looking for an experienced Senior Data Scientist to help us benchmark against industry standards to uncover insights to improve and optimize the long term supply chain strategy for Amazons Sub-Same-Day business.
US, NY, New York
Job summaryJob summaryAmazon is looking for a passionate, talented, and inventive Applied Scientist with a strong machine learning background to help build industry-leading language technology.Our mission is to provide a delightful experience to Amazon’s customers by pushing the envelope in Natural Language Processing (NLP), Natural Language Understanding (NLU), Dialog management, conversational AI and Machine Learning (ML).As part of our AI team in Amazon AWS, you will work alongside internationally recognized experts to develop novel algorithms and modeling techniques to advance the state-of-the-art in human language technology. Your work will directly impact millions of our customers in the form of products and services, as well as contributing to the wider research community. You will gain hands on experience with Amazon’s heterogeneous text and structured data sources, and large-scale computing resources to accelerate advances in language understanding.We are hiring primarily in Conversational AI / Dialog System Development areas: NLP, NLU, Dialog Management, NLG.This role can be based in NYC, Seattle or Palo Alto.Inclusive Team CultureHere at AWS, we embrace our differences. We are committed to furthering our culture of inclusion. We have ten employee-led affinity groups, reaching 40,000 employees in over 190 chapters globally. We have innovative benefit offerings, and host annual and ongoing learning experiences, including our Conversations on Race and Ethnicity (CORE) and AmazeCon (gender diversity) conferences.Work/Life BalanceOur team puts a high value on work-life balance. It isn’t about how many hours you spend at home or at work; it’s about the flow you establish that brings energy to both parts of your life. We believe striking the right balance between your personal and professional life is critical to life-long happiness and fulfillment. We offer flexibility in working hours and encourage you to find your own balance between your work and personal lives.Mentorship & Career GrowthOur team is dedicated to supporting new members. We have a broad mix of experience levels and tenures, and we’re building an environment that celebrates knowledge sharing and mentorship. Our senior members enjoy one-on-one mentoring and thorough, but kind, code reviews. We care about your career growth and strive to assign projects based on what will help each team member develop into a better-rounded engineer and enable them to take on more complex tasks in the future.
US, VA, Arlington
Job summaryAmazon is investing heavily in building a world class advertising business and we are responsible for defining and delivering a collection of self-service performance advertising products that drive discovery and sales. Our products are strategically important to our Retail and Marketplace businesses driving long term growth. We deliver billions of ad impressions and millions of clicks daily and are breaking fresh ground to create world-class products. We are highly motivated, collaborative and fun-loving with an entrepreneurial spirit and bias for action. With a broad mandate to experiment and innovate, we are growing at an unprecedented rate with a seemingly endless range of new opportunities.Sponsored Products helps merchants, retail vendors, and brand owners succeed via native advertising that grows incremental sales of their products sold through Amazon. The Sponsored Products Ad Marketplace organization optimizes the systems and ad placements to match advertiser demand with publisher supply using a combination of machine learning, big data analytics, ultra-low latency high-volume engineering systems, and quantitative product focus. Our goals are to help buyers discover new products they love, be the most efficient way for advertisers to meet their business objectives, and to build a major, sustainable business that helps Amazon continuously innovate on behalf of all customers.We are seeking a Sr. Applied Science Manager who has a solid background in applied Machine Learning and AI, deep passion for building data-driven products, ability to communicate data insights and scientific vision, and has a proven track record of leading both applied scientists and software engineers to execute complex projects and deliver business impacts.In this team, Machine Learning and Deep Learning technologies including Semantic Retrieval, Natural Language Processing (NLP), Information Extraction, Image Understanding, Learning to Rank are used to match shoppers' search queries to ads with per impression prediction models that run in real-time with tight latency budgets. Models are trained using self-supervised techniques, transfer learning, and supervised training using labeled datasets. Knowledge distillation and model compression techniques are used to optimize model performance for production serving.The Senior Manager role will lead science and engineering efforts in these areas for Amazon Search pages WW. The person in this role is responsible for: maintaining the consistent and long term reliability for the models and the delivery services that power them, managing diverse teams across multiple domains, and collaborating cross-functional with other senior decision makers. Our critical LPs for this role are Think Big, Are Right A lot, and Earns Trust. What is key is that the leader will need a dynamic mindset to build systems that are flexible and will scale.In this role, you will:· Lead a group of both applied scientists and software engineers to deliver machine-learning and AI solutions to production.· Advance team's engineering craftsmanship and drive continued scientific innovation as a thought leader and practitioner.· Develop science and engineering roadmap, run Sprint/quarter and annual planning, and foster cross-team collaboration to execute complex projects.· Perform hands-on data analysis, build machine-learning models, run regular A/B tests, and communicate the impact to senior management.· Hire and develop top talents, provide technical and career development guidance to both scientists and engineers in the organization.Locations: Seattle, WA; New York, NY; Arlington, VA
US, WA, Seattle
Job summaryWorkforce Staffing (WFS) brings together the workforce powering Amazon’s ability to delight customers: the Amazon Associate. With over 1M hires, WFS supports sourcing, hiring, and developing the best talent to work in our fulfillment centers, sortation centers, delivery stations, shopping sites, Prime Air locations, and more.WFS' Funnel Science and Analytics team is looking for a Research Scientist. This individual will be responsible for conducting experiments and evaluating the impact of interventions when conducting experiments is not feasible. The perfect candidate will have the applied experience and the theoretical knowledge of policy evaluation and conducting field studies.Key job responsibilitiesAs a Research Scientist (RS), you will do causal inference, design studies and experiments, leverage data science workflows, build predictive models, conduct simulations, create visualizations, and influence science and analytics practice across the organization.Provide insights by analyzing historical data from databases (Redshift, SQL Server, Oracle DW, and Salesforce).Identify useful research avenues for increasing candidate conversion, test, and create well written documents to communicate to technical and non-technical audiences.About the teamFunnel Science and Analytics team finds ways to maximize the conversion and early retention of every candidate who wants to be an Amazon Associate. By focusing on our candidates, we improve candidate and business outcomes, and Amazon takes a step closer to being Earth’s Best Employer.
US, NY, New York
Job summaryAmazon is looking for a passionate, talented, and inventive Applied Scientist with a strong machine learning background to help build industry-leading language technology.Our mission is to provide a delightful experience to Amazon’s customers by pushing the envelope in Natural Language Processing (NLP), Natural Language Understanding (NLU), Dialog management, conversational AI and Machine Learning (ML).As part of our AI team in Amazon AWS, you will work alongside internationally recognized experts to develop novel algorithms and modeling techniques to advance the state-of-the-art in human language technology. Your work will directly impact millions of our customers in the form of products and services, as well as contributing to the wider research community. You will gain hands on experience with Amazon’s heterogeneous text and structured data sources, and large-scale computing resources to accelerate advances in language understanding.We are hiring primarily in Conversational AI / Dialog System Development areas: NLP, NLU, Dialog Management, NLG.This role can be based in NYC, Seattle or Palo Alto.Inclusive Team CultureHere at AWS, we embrace our differences. We are committed to furthering our culture of inclusion. We have ten employee-led affinity groups, reaching 40,000 employees in over 190 chapters globally. We have innovative benefit offerings, and host annual and ongoing learning experiences, including our Conversations on Race and Ethnicity (CORE) and AmazeCon (gender diversity) conferences.Work/Life BalanceOur team puts a high value on work-life balance. It isn’t about how many hours you spend at home or at work; it’s about the flow you establish that brings energy to both parts of your life. We believe striking the right balance between your personal and professional life is critical to life-long happiness and fulfillment. We offer flexibility in working hours and encourage you to find your own balance between your work and personal lives.Mentorship & Career GrowthOur team is dedicated to supporting new members. We have a broad mix of experience levels and tenures, and we’re building an environment that celebrates knowledge sharing and mentorship. Our senior members enjoy one-on-one mentoring and thorough, but kind, code reviews. We care about your career growth and strive to assign projects based on what will help each team member develop into a better-rounded engineer and enable them to take on more complex tasks in the future.
US, WA, Seattle
Job summaryAmazon's Weblab team enables experimentation at massive scale to help Amazon build better products for customers. A/B testing is in Amazon's DNA and we're at the core of how Amazon innovates on behalf of customers. We are seeking a skilled Applied Scientist to help us build the future of experimentation systems at Amazon.About you:You have an entrepreneurial spirit and want to make a big impact on Amazon and its customers. You are excited about cutting-edge research on unsupervised learning, graph algorithms, and causal inference in the intersection between Machine Learning, Statistics, and Econometrics. You enjoy building massive scale and high performance systems but also have a bias for delivering simple solutions to complex problems. You're looking for a career where you'll be able to build, to deliver, and to impress. You challenge yourself and others to come up with better solutions. You develop strong working relationships and thrive in a collaborative team environment.About us together:We're going to help Amazon make better long term decisions by designing and delivering A/B-testing systems for long-term experiments, and by using these systems to figure out how near term behavior impacts long term growth and profitability. Our work will inform some of the biggest decisions at Amazon. Along the way, we're going to face seemingly insurmountable challenges. We're going to argue about how to solve them, and we'll work together to find a solution that is better than each of the proposals we came in with. We'll make tough decisions, but we'll all understand why. We'll be the dream team.We have decades of combined experience on the team in many areas science and engineering so it's a great environment in which to learn and grow. A/B testing is one of the hottest areas of research and development in the world today and this is a chance to learn how it works in the company known for pioneering its use.
US, CA, Santa Clara
Job summaryAWS AI/ML is looking for world class scientists and engineers to join its AI Research and Education group working on building automated ML solutions for planetary-scale sustainability and geospatial applications. Our team's mission is to develop ready-to-use and automated solutions that solve important sustainability and geospatial problems. We live in a time wherein geospatial data, such as climate, agricultural crop yield, weather, landcover, etc., has become ubiquitous. Cloud computing has made it easy to gather and process the data that describes the earth system and are generated by satellites, mobile devices, and IoT devices. Our vision is to bring the best ML/AI algorithms to solve practical environmental and sustainability-related R&D problems at scale. Building these solutions require a solid foundation in machine learning infrastructure and deep learning technologies. The team specializes in developing popular open source software libraries like AutoGluon, GluonCV, GluonNLP, DGL, Apache/MXNet (incubating). Our strategy is to bring the best of ML based automation to the geospatial and sustainability area.We are seeking an experienced Applied Scientist for the team. This is a role that combines science knowledge (around machine learning, computer vision, earth science), technical strength, and product focus. It will be your job to develop ML system and solutions and work closely with the engineering team to ship them to our customers. You will interact closely with our customers and with the academic and research communities. You will be at the heart of a growing and exciting focus area for AWS and work with other acclaimed engineers and world famous scientists. You are also expected to work closely with other applied scientists and demonstrate Amazon Leadership Principles (https://www.amazon.jobs/en/principles).Strong technical skills and experience with machine learning and computer vision are required. Experience working with earth science, mapping, and geospatial data is a plus. Our customers are extremely technical and the solutions we build for them are strongly coupled to technical feasibility.About the teamInclusive Team CultureAt AWS, we embrace our differences. We are committed to furthering our culture of inclusion. We have ten employee-led affinity groups, reaching 40,000 employees in over 190 chapters globally. We have innovative benefit offerings, and host annual and ongoing learning experiences, including our Conversations on Race and Ethnicity (CORE) and AmazeCon (gender diversity) conferences. Amazon’s culture of inclusion is reinforced within our 14 Leadership Principles, which remind team members to seek diverse perspectives, learn and be curious, and earn trust.Work/Life BalanceOur team puts a high value on work-life balance. It isn’t about how many hours you spend at home or at work; it’s about the flow you establish that brings energy to both parts of your life. We believe striking the right balance between your personal and professional life is critical to life-long happiness and fulfillment. We offer flexibility in working hours and encourage you to find your own balance between your work and personal lives.Mentorship & Career GrowthOur team is dedicated to supporting new members. We have a broad mix of experience levels and tenures, and we’re building an environment that celebrates knowledge sharing and mentorship. Our senior members enjoy one-on-one mentoring. We care about your career growth and strive to assign projects based on what will help each team member develop into a better-rounded scientist and enable them to take on more complex tasks in the future.Interested in this role? Reach out to the recruiting team with questions or apply directly via amazon.jobs.
US, CA, Santa Clara
Job summaryAWS AI/ML is looking for world class scientists and engineers to join its AI Research and Education group working on building automated ML solutions for planetary-scale sustainability and geospatial applications. Our team's mission is to develop ready-to-use and automated solutions that solve important sustainability and geospatial problems. We live in a time wherein geospatial data, such as climate, agricultural crop yield, weather, landcover, etc., has become ubiquitous. Cloud computing has made it easy to gather and process the data that describes the earth system and are generated by satellites, mobile devices, and IoT devices. Our vision is to bring the best ML/AI algorithms to solve practical environmental and sustainability-related R&D problems at scale. Building these solutions require a solid foundation in machine learning infrastructure and deep learning technologies. The team specializes in developing popular open source software libraries like AutoGluon, GluonCV, GluonNLP, DGL, Apache/MXNet (incubating). Our strategy is to bring the best of ML based automation to the geospatial and sustainability area.We are seeking an experienced Applied Scientist for the team. This is a role that combines science knowledge (around machine learning, computer vision, earth science), technical strength, and product focus. It will be your job to develop ML system and solutions and work closely with the engineering team to ship them to our customers. You will interact closely with our customers and with the academic and research communities. You will be at the heart of a growing and exciting focus area for AWS and work with other acclaimed engineers and world famous scientists. You are also expected to work closely with other applied scientists and demonstrate Amazon Leadership Principles (https://www.amazon.jobs/en/principles).Strong technical skills and experience with machine learning and computer vision are required. Experience working with earth science, mapping, and geospatial data is a plus. Our customers are extremely technical and the solutions we build for them are strongly coupled to technical feasibility.About the teamInclusive Team CultureAt AWS, we embrace our differences. We are committed to furthering our culture of inclusion. We have ten employee-led affinity groups, reaching 40,000 employees in over 190 chapters globally. We have innovative benefit offerings, and host annual and ongoing learning experiences, including our Conversations on Race and Ethnicity (CORE) and AmazeCon (gender diversity) conferences. Amazon’s culture of inclusion is reinforced within our 14 Leadership Principles, which remind team members to seek diverse perspectives, learn and be curious, and earn trust.Work/Life BalanceOur team puts a high value on work-life balance. It isn’t about how many hours you spend at home or at work; it’s about the flow you establish that brings energy to both parts of your life. We believe striking the right balance between your personal and professional life is critical to life-long happiness and fulfillment. We offer flexibility in working hours and encourage you to find your own balance between your work and personal lives.Mentorship & Career GrowthOur team is dedicated to supporting new members. We have a broad mix of experience levels and tenures, and we’re building an environment that celebrates knowledge sharing and mentorship. Our senior members enjoy one-on-one mentoring. We care about your career growth and strive to assign projects based on what will help each team member develop into a better-rounded scientist and enable them to take on more complex tasks in the future.Interested in this role? Reach out to the recruiting team with questions or apply directly via amazon.jobs.