WeAreDevelopers Logo WeAreDevelopers
Sign in
For You LIVE
WeAreDevelopers LIVE • Jun 12, 2020

Ranking Amazon Reviews by Quality with Pointwise Ratings learned from Pairwise Data

Tanmay Bakshi

How can you accurately rank subjective customer reviews without explicit quality labels? Discover how combining BERT and Siamese networks for pairwise classification solves this classic NLP limitation.

Pause
Mute Enter Fullscreen
#1 about 3 min

Formulating the challenge of quantifying Amazon review helpfulness

Translating qualitative text feedback into a single continuous value enables programmatic sorting of reviews.

#2 about 3 min

Identifying core features of highly useful customer reviews

Evaluating structural length and information density unpacks how models evaluate text without overfitting to arbitrary bounds.

#3 about 3 min

Leveraging meta ratings to construct AI training datasets

Harnessing community moderation reactions provides reliable data labels for model training without manual string annotation.

#4 about 7 min

Linearly separating input features via deep embedding spaces

Projecting complex multimodal inputs into mathematically searchable vector spaces streamlines linear classification logic.

#5 about 3 min

Extracting structural word meanings using semantic vector representations

Mapping sequential words into dense semantic representations enables direct arithmetic manipulation of natural language elements.

#6 about 4 min

Processing sequential text data with recurrent neural networks

Analyzing token sequences chronologically mimics time series evaluations for foundational context derivation.

#7 about 5 min

Addressing operational limitations of recurrent neural network architectures

Analyzing texts sequentially creates severe processing bottlenecks and forces models to endure domain-specific retraining cycles.

#8 about 5 min

Parallelizing language comprehension with transformers and BERT models

Bypassing sequential processing limits allows neural networks to evaluate complete textual contexts simultaneously.

#9 about 3 min

Training contextual language models through masked word prediction

Masking contextual segments of input sentences forces models to deduce vocabulary mappings bidirectionally.

#10 about 1 min

Minimizing vocabulary limitations using WordPiece text tokenization techniques

Breaking unknown words into strict root components completely eliminates failed out-of-vocabulary inputs during text analysis.

#11 about 7 min

Modeling character-level predictions with self attention network logic

Demonstrating self-attention mechanics through a custom character-prediction network reinforces structural language comprehension capabilities.

#12 about 3 min

Handling dataset uncertainty using Laplace rule of succession

Modifying raw upvote fractions algorithmically mitigates natural variance in user engagement rates across product ratings.

#13 about 3 min

Extracting scalar metrics from pairwise text classification routines

Passing comparative data permutations into unified networks generates granular ranking scores instead of basic binary classifications.

#14 about 3 min

Evaluating continuous value review rankings during live inference

Executing the trained ranking model on real-world text inputs highlights the rapid inference speeds of robust architectures.

Matching moments

2:18 min

Major breakthroughs shaping the artificial intelligence landscape

Nico Axtmann · WWC 2022

1:52 min

Modern natural language processing using Transformers and deep learning

Jan Schweiger · WWC 2022

2:33 min

Evaluating AI comprehension and output quality

Perf + AI

2:23 min

Historical breakthroughs in natural language processing models

Mary Grygleski Mary Grygleski · LIVE

2:36 min

Exploring high-level Python frameworks for accelerated enterprise artificial intelligence

Paul Graham Paul Graham · LIVE

1:59 min

Exploring the AI technology stack for chatbot demos

Akmal Chaudhri Akmal Chaudhri · LIVE

More from WeAreDevelopers LIVE

Aliens in Quarantine, Python Deprecates Loops, and AI Buys Old Books - Zbyszek Tenerowicz thumbnail
5:43
5 days ago

Aliens in Quarantine, Python Deprecates Loops, and AI Buys Old Books - Zbyszek Tenerowicz

Chris Heilmann, Daniel Cranney

WeAreDevelopers LIVE - Node and Package Security thumbnail
1:08:30
5 days ago

WeAreDevelopers LIVE - Node and Package Security

Chris Heilmann, Daniel Cranney, Zbyszek Tenerowicz

WeAreDevelopers LIVE - Streaming HTML thumbnail
1:10:26
12 days ago

WeAreDevelopers LIVE - Streaming HTML

Chris Heilmann, Daniel Cranney, Jad Joubran

WeAreDevelopers LIVE - CODE100 at WeAreDevelopers Europe thumbnail
1:02:34
19 days ago

WeAreDevelopers LIVE - CODE100 at WeAreDevelopers Europe

Chris Heilmann, Daniel Cranney, Justin Halsall, Werner Jeschko, Nicole Jeschko, Mirco Tomić

WeAreDevelopers LIVE - Web Performance in an AI World thumbnail
1:07:59
26 days ago

WeAreDevelopers LIVE - Web Performance in an AI World

Chris Heilmann, Daniel Cranney

FIFA Hack, Midjourney Does Ultrasounds, and Trampoline Traffic Lights - Nicole Jeschko and Werner Jeschko thumbnail
5:55
about 2 months ago

FIFA Hack, Midjourney Does Ultrasounds, and Trampoline Traffic Lights - Nicole Jeschko and Werner Jeschko

Chris Heilmann, Daniel Cranney

Upcoming sessions on this topic

Open session

World Congress 2026 North America

Evals Are Infra: Building AI Systems Developers Can Actually Trust

Phoebe Wang

Member of Technical Staff at OpenAI

Phoebe Wang
Open session

World Congress 2026 North America

You Can’t Re-Run Sunlight: Designing ML Data Architectures for Physical AI

An Phan

Senior Data Infrastructure Engineer @ Hippo Harvest

An Phan
Open session

World Congress 2026 North America

Building Pragmatic AI: 10 AI Features Your Users Actually Want

Jonathan "J." Tower

.NET Foundation Board | 12x Microsoft MVP | Founder & Consultant

Jonathan "J." Tower
Open session

World Congress 2026 North America

No Single Model to Rule Them All: Building Resilient AI Agents Across Open & Closed LLMs

Emmanuel Acheampong

Senior Manager Developer Relations at Crusoe AI

Emmanuel Acheampong
Open session

World Congress 2026 North America

Proactive AI That Doesn’t Annoy Users: Building Context-Aware Notification Systems

Raju Dandigam Dandigam

Engineering Manager at Navan

Raju Dandigam Dandigam
Open session

World Congress 2026 North America

Headroom: A Context Optimization Layer for LLM Applications

Tejas Chopra

Senior Software Engineer at Netflix

Tejas Chopra

Related videos

A beginner’s guide to modern natural language processing thumbnail
56:46
LIVE • over 2 years ago

A beginner’s guide to modern natural language processing

Jodie Burchell

Hybrid AI: Next Generation Natural Language Processing thumbnail
21:01
WWC 2022 • about 4 years ago

Hybrid AI: Next Generation Natural Language Processing

Jan Schweiger

How AI Models Get Smarter thumbnail
35:16
WWC 2025 • about 1 year ago

How AI Models Get Smarter

Ankit Patel

How We Built a Machine Learning-Based Recommendation System (And Survived to Tell the Tale) thumbnail
27:23
WWC 2023 • almost 3 years ago

How We Built a Machine Learning-Based Recommendation System (And Survived to Tell the Tale)

Dora Petrella

Evaluating AI models for code comprehension thumbnail
22:03
WWC 2025 • 12 months ago

Evaluating AI models for code comprehension

Merrill Lutsky

Inside the Mind of an LLM thumbnail
27:11
WWC 2025 • 12 months ago

Inside the Mind of an LLM

Emanuele Fabbiani

Related articles

From learning to earning

Jobs that call for the skills explored in this talk.