The article outlines JD’s evolution from rule‑based product suggestions in 2012 to a sophisticated, AI‑driven, multi‑screen personalized recommendation platform, detailing its product types, system architecture, data collection, offline and online computation, and the core recommendation engine that powers features like “Guess You Like.”
The article details JD.com’s journey from rule‑based product recommendations in 2012 to a sophisticated, AI‑driven personalized recommendation system, describing its multi‑screen product types, data collection, offline and online computation pipelines, and the modular architecture of its recommendation engine.
This article details the design, architecture, and AI‑driven features of 58’s intelligent monitoring platform, explaining how multi‑dimensional data collection, predictive analytics, and smart alarm merging ensure continuous, automated observability across network, server, application, and business layers.
This article summarizes Hulu Chief Research Officer Xie Xiaohui’s presentation on why video content understanding is essential, the technical challenges involved, and Hulu’s end‑to‑end solutions—including fine‑grained segmentation, logo and subtitle detection, automated pipelines, tagging taxonomy, content generation, and vector embeddings—to improve recommendation, advertising, and search for massive video libraries.
This article introduces recommendation ranking, breaks down its components, explains how to build ranking models, and provides additional insights, while also presenting the author's background and related job opportunities in the e‑commerce recommendation field.
This article presents a comprehensive solution to the absolute semantic recognition competition, detailing the problem background, dataset, evaluation metrics, feature engineering, model architecture—including Attention, Capsule, Bi‑GRU, and BERT—and analysis of results and lessons learned.
Learn how to overcome the complexities of machine‑learning workflow management by installing Kubeflow Pipelines on Alibaba Cloud using Kustomize, including prerequisites, TLS setup, persistent storage with SSD disks, image registry replacement, and deployment verification steps.
The article details a multi‑turn dialogue intent classification pipeline that extracts and expands labeled utterances, preprocesses text with custom tokenization, trains a two‑layer CNN‑Highway and a multi‑head self‑attention model, analyzes errors, and achieves up to 98.7% accuracy on a large, balanced dataset.
The article traces the historical development of recommendation systems from early manual and hot‑ranking methods through natural ranking and machine‑learning‑based scoring, discusses the Matthew effect and its mitigation via randomization, multi‑objective weighting, and pipeline architectures, and outlines modern personalization and recall strategies for e‑commerce platforms.
This article explores how artificial intelligence is influencing frontend development, detailing recent breakthroughs, practical frameworks like TensorFlow.js, real‑world applications, advantages, limitations, and emerging solutions for deploying machine learning models directly in browsers.
By replacing rule‑based pipelines with an active‑learning‑driven random‑forest model that extracts clustering, signal, association, IP, and temporal features, Gaode accurately identifies mobile, cloned, and moved Wi‑Fi, cutting large‑error network‑positioning cases by ~18% and boosting overall positioning precision.
This article presents a comprehensive overview of Alibaba Retail's B2B2C intelligent recommendation platform, detailing its business context, three core recommendation scenarios, system architecture, matching algorithms such as item‑CF, graph embedding and user‑CF, as well as the evolution of ranking models and feature engineering practices.
This article systematically reviews the evolution of attention mechanisms—from early additive and multiplicative forms to self‑attention and multi‑head variants—explaining their core three‑step framework, key differences, and why they have become essential across NLP, vision, and broader AI applications.
This article presents a case study on building and deploying a lead‑quality classification model for a high‑value, low‑frequency real‑estate platform, covering business context, data challenges, sampling strategies, feature engineering, model selection, tuning, evaluation metrics, interpretability analysis, and observed performance improvements.
This article dispels five widespread AI misconceptions—from believing AI works like the human brain to thinking it is bias‑free—while offering practical guidance on recognizing AI limits, improving data quality, managing risks, and applying AI responsibly across industries.
This article explains how sequence embedding can be applied to user browsing data on an automotive website to predict intended car models, addressing data sparsity and leveraging temporal information to improve prediction accuracy by 3%.
The article explains how recommendation systems balance exploitation and exploration, introduces diversity metrics such as temporal, spatial, and coverage, and presents a determinantal point process (DPP) based algorithm accelerated by Cholesky decomposition and greedy inference, demonstrating significant speedups and improved relevance‑diversity trade‑offs in experiments.
Magic Mirror is a big‑data‑based visual analytics platform that lowers the barrier of machine‑learning for non‑experts while accelerating expert workflows through visual UI, modular algorithms, distributed feature generation, and automated binary‑classification modeling.
This article presents the design, implementation, and evaluation of X‑monitor, an AI‑driven, adaptive, multi‑modal financial data quality monitoring platform that combines rule‑based and self‑learning strategies to improve detection efficiency, accuracy, and flexibility for large‑scale securities‑firm data streams.
This article describes how Tujia applied embedding techniques, particularly a Skip‑Gram model, to build an item‑to‑item similarity recommender for low‑frequency, highly personalized homestay listings, detailing the data preparation, model architecture, training process, evaluation results, practical improvements, and future directions.
This article explains how Alibaba's App recommendation pipeline integrates marketing scenario cards using weak personalization and machine‑learning models, detailing the metrics, feature engineering, recall and ranking strategies that together raise exposure revenue and click‑through performance.
To overcome the limitations of handcrafted rules for binding users’ reported start locations to the correct road segment, Amap built a data‑driven, list‑wise learning‑to‑rank model that leverages real‑travel and planning data, achieving a 10 % error reduction and 40 % accuracy gain on difficult origin‑snapping cases.
This article explains how threat intelligence is generated by defining it as judged security information, outlines methods for collecting and evaluating security data, introduces a two‑dimensional reliability/quality rating system, and provides a step‑by‑step engineering workflow for enterprise threat‑intelligence operations.
iQIYI’s ZoomAI combines super‑resolution, denoising, sharpening and color‑enhancement AI models to restore low‑quality historic Chinese films, achieving up to 500‑fold faster processing than manual methods while preserving original visual characteristics.
This article explains why and how to discretize continuous features for categorical targets, compares three common discretization strategies, and details a Bayesian optimal method (MODL) with its mathematical formulation and implementation references.
The article compares traditional credit‑scoring‑card models with modern machine‑learning approaches for financial risk control, detailing feature selection criteria, non‑linear handling, data characteristics, practical ML techniques, large‑scale modeling challenges, and summarizing insights for future development.
This article presents a curated roundup of the most starred GitHub repositories in May, covering terminals, Python learning resources, algorithm collections, command‑line guides, Java tutorials, and a machine‑learning course, each with brief descriptions and star counts.
This article presents detailed study notes on the book “Computational Advertising”, covering an overview, ad system architecture, targeted advertising, online advertising, the E‑E algorithm, and additional insights, accompanied by illustrative diagrams to aid understanding of modern advertising technologies.
The article summarizes the 58 Group technical salon where experts presented the microservice‑based recommendation system architecture, data and strategy layers, and the internally built search ranking platform covering sampling, feature engineering, and model training, highlighting practical implementations and lessons learned.
The third Sohu Content Recognition Algorithm Competition wrapped up after 65 days, drawing 218 universities and 982 teams, showcasing over 5,000 submissions, awarding prizes to top university teams, and sharing key lessons on data focus, model fusion, and engineering practice for AI development.
This article provides a detailed, English-language summary of key statistical learning concepts—including perceptron, k‑nearest neighbors, Naive Bayes, decision trees, logistic regression, support vector machines, boosting, EM, HMM, neural networks, K‑Means, bagging, Apriori and dimensionality reduction—complete with formulas, algorithm steps, and illustrative diagrams to aid interview preparation.
This article presents an in‑depth overview of Alibaba's XiaoMi chatbot, covering its evolution from traditional customer service to an intelligent service platform, the design of emotion recognition and response models, customer emotion soothing pipelines, service quality detection, generative emotional dialogue, and future work on session satisfaction estimation.
At QCon Beijing 2019, four Beike technology experts presented the practical use and challenges of machine learning for user profiling, deep‑learning‑based house‑quality scoring, intelligent customer‑service systems, and AI‑driven floor‑plan generation, summarizing the architecture, data pipelines, model evolution, and future improvement directions.
This comprehensive guide explores a wide range of anomaly detection techniques—including time‑series methods, statistical models, distance‑based approaches, tree‑based isolation forests, graph algorithms, behavior‑sequence Markov models, and supervised machine‑learning models—detailing their principles, formulas, and practical scenarios for detecting outliers in advertising, fraud, and system monitoring.
This article provides a comprehensive overview of deep learning, covering essential mathematics and machine learning fundamentals, modern deep network architectures and regularization techniques, advanced research topics such as structured probabilistic models and generative methods, and a curated reading list for practitioners.
This guide walks you through getting started with Google Colab, covering setup, basic notebook usage, useful configurations, Google Drive mounting, and how the platform supports machine‑learning teaching and GPU acceleration, all without any local installation.
Using AI to automatically understand and secure audio content, this article discusses the challenges of manual audio analysis, outlines a four‑step pipeline—audio segmentation, speech‑to‑text, labeling, and synthesis—and describes models such as VAD, ASR, sound classification, text recognition, and behavior detection for live‑stream moderation.
This article presents a comprehensive overview of financial risk management, detailing business challenges such as identity verification and fraud, and describing technical solutions including feature engineering, sample handling, model optimization, and online validation, emphasizing the integration of data-driven AI techniques throughout the process.
SQLFlow is an open‑source project that connects diverse SQL engines (MySQL, Hive, SparkSQL, etc.) with AI frameworks (TensorFlow, PyTorch, XGBoost, etc.) through extended SQL syntax, enabling analysts to train and predict models using only a few SQL statements while aiming for high scalability and performance.
The article details the Travel team's award‑winning solution for the WSDM Cup 2019 fake‑news detection task, describing data analysis, preprocessing, label‑propagation augmentation, a BERT‑based baseline, a three‑stage multi‑level model‑fusion framework, experimental results, and future directions.
This article examines the differences and similarities between classification (pointwise) and ranking (pairwise) models for recommendation systems, covering their probabilistic foundations, loss functions, parameter updates, and practical implications such as sensitivity to statistical features and robustness.
This article examines the rise of RDMA high‑performance networking, explains its technical advantages over traditional TCP/IP, showcases real‑world deployments in machine learning at Baidu, and explores future use cases in storage, GPU communication, and core services.
Learn how to create two Python chatbots—one using the wxpy library with the Turing API and another trained with ChatterBot—by following concise step‑by‑step instructions and sample code, enabling you to build functional conversational agents with minimal effort.
This article shares the experience of Xu An, runner‑up in the 2019 Tencent Advertising Algorithm Competition, detailing practical advice on feature engineering, model selection, efficiency tricks, personal habits, contest rhythm, and learning resources for aspiring participants.
This article explains how personalized recommendation systems work in e‑commerce platforms like Taobao, covering the fundamentals of machine learning, feature engineering, model building, evaluation, and the business motivations behind delivering unique, user‑specific product suggestions.
This comprehensive guide walks you through the fundamentals of speech recognition, explains how it works, compares Python packages, shows step‑by‑step installation of SpeechRecognition, demonstrates processing audio files and live microphone input, and offers techniques for handling noise and multilingual transcription.
This article examines the role of search as a user guidance channel in e-commerce, outlining product requirements, user flow stages, and various algorithmic solutions—including query understanding, suggestion, rewriting, retrieval, and ranking—while also comparing implementations across major Chinese platforms.
This report details Ctrip Financial's end‑to‑end risk control development, covering business overview, a three‑layer data platform, the progression of credit scoring and anti‑fraud models from rule‑based to advanced AI techniques, and the evaluation, monitoring, and social‑network‑based fraud detection strategies employed.
This article introduces the AdaBoost algorithm, explains its boosting principle, describes the AdaBoostClassifier and AdaBoostRegressor classes in scikit‑learn, provides a complete Python example with data loading, model training, prediction, evaluation, and visualisation, and discusses the algorithm’s advantages, disadvantages, and detailed iterative process.
This article provides a concise overview of ten fundamental machine learning algorithms—linear regression, logistic regression, linear discriminant analysis, naive Bayes, K‑nearest neighbors, learning vector quantization, support vector machines, decision trees, bagging/random forest, and boosting/AdaBoost—explaining their principles, typical use cases, and key characteristics.
The study introduces a voting ensemble of Random Forest, AdaBoost, and XGBoost to predict hot‑migration success in cloud environments, achieving 97.44% accuracy and cutting timeout failures by roughly 80%, while quantifying feature importance—primarily CPU, network traffic, and memory—to guide proactive resource allocation.
A champion shares insights and strategies from the Tencent Ad Algorithm Competition, focusing on machine learning models, feature engineering, and time-series approaches for the 2019 edition's ad exposure prediction challenge.
The article presents a systematic, top‑down methodology for machine‑learning performance optimization—covering principles, benchmark‑driven loops, foundational hardware and software checks, profiling tools, throughput and latency metrics, and practical techniques for IO, compute, mixed‑precision, and distributed training to maximize resource utilization.
Alibaba’s senior engineers share their curated list of essential technical books—from software testing and design patterns to AI, machine learning, reinforcement learning, Rust programming, and database architecture—explaining why each title is valuable for developers seeking deeper knowledge and practical insights.
This article explains the fundamental principle of gradient descent as the steepest‑descent optimization method, derives its direction using Taylor expansion and the Cauchy‑Schwarz inequality, illustrates why it can be slow on functions like Rosenbrock, and discusses its advantages and convergence properties.
This article explains how to use TensorBoard and the Timeline tool to monitor TensorFlow GPU utilization, identify operation bottlenecks, and visualize execution times, including code examples and steps for exporting and merging profiling data for repeated runs.
This article explains the design of a live‑streaming recommendation system, covering its overall architecture, ranking, content‑based and collaborative‑filtering methods, similarity calculations, multi‑algorithm fusion, sorting, user profiling, and evaluation metrics with practical examples and diagrams.
This article introduces meta‑learning (learning to learn), its historical roots, explains why it excels in small‑sample and multi‑task settings, contrasts it with supervised and reinforcement learning, and outlines the theoretical reasons it enables rapid few‑shot adaptation.
This article describes a monitoring system that leverages machine learning to perform threshold‑free, real‑time anomaly detection on macro business indicators such as network traffic and access volume, detailing its architecture, sample labeling, model training, and multi‑level alarm strategies.
This article compiles a comprehensive set of beginner-friendly machine‑learning resources—including recommended programming languages, essential textbooks, video courses, influential blogs, competition platforms, and notable conference papers—to help newcomers build a solid foundation and practical experience.
This article introduces how machine learning problems are formulated as optimization tasks, explains the construction of objective functions with examples such as linear regression, robust fitting, regularization, and demonstrates various applications ranging from K‑means clustering to image inpainting and 3D reconstruction.
This article explores the evolution of Generative Adversarial Networks' objective functions, detailing the shift from Jensen‑Shannon divergence to f‑divergence, IPM‑based approaches, and auxiliary losses, while highlighting their impact on stability and performance across image, audio, and text generation tasks.
This article reviews the evolution of Taobao’s search and recommendation technology, tracing its journey from simple statistical models and rule‑based systems through large‑scale machine learning and real‑time online learning to modern deep‑learning and cognitive intelligence approaches that drive e‑commerce innovation.
This article explores how Field-aware Factorization Machines (FFM) can be employed to replace multi‑path recall strategies in industrial recommendation systems, detailing model principles, embedding construction, integration of user, item and context features, performance considerations, and potential for unifying recall and ranking stages.
This article introduces Variational Autoencoders (VAEs), compares them with GANs, explains the underlying variational inference principle, and details how VAEs are trained using the evidence lower bound, complemented by visual diagrams and key equations.
This comprehensive guide explores anomaly (outlier) detection across diverse methods—including time‑series analysis, statistical tests, distance metrics, matrix factorization, graph approaches, behavior‑sequence modeling, and supervised machine‑learning models—highlighting their principles, formulas, and practical use cases such as fraud prevention and system monitoring.
This article outlines Alibaba Retail's demand forecasting workflow, describing the evolution from simple mean and time‑series models to machine‑learning and deep‑learning approaches, the incorporation of feature engineering, operational plans, and methods for estimating prediction uncertainty to support intelligent replenishment.
This article presents a comprehensive overview of personalized recommendation systems, covering their purpose, common algorithms, development challenges, the multi‑layer architecture used at DataGrand, optimization techniques, and the range of services offered to enterprise customers.
AI brings convenience but also new security challenges; this article explains the two main sources of AI safety issues, details adversarial example techniques, showcases applications such as face‑recognition attacks and robust captcha designs, and highlights Alibaba’s research and the IJCAI‑19 AI adversarial competition.
This comprehensive guide introduces data mining and machine learning concepts, walks through Python data preprocessing techniques, reviews common classification algorithms, demonstrates an Iris flower classification case, and offers practical tips for selecting the most suitable algorithm for a given problem.
This article introduces the open‑source H2O platform, explains how to install and use its Python API for data loading, preprocessing, model training with GBM and AutoML, evaluates results with AUC, and describes model deployment via POJO/MOJO as well as the visual Flow UI, concluding with reflections on the role of automated modeling in data science.
The article explores how Alibaba’s DAMO Academy uses AI to detect and dismantle online misinformation through a three‑step analysis of source credibility, content verification, and propagation paths, highlighting a record‑breaking 81% accuracy in the SemEval fake‑news competition.
Vipshop's in‑house sentiment monitoring platform integrates web‑scraped reviews, WeChat comments and internal service messages, applying lexical sentiment scoring, dictionary‑based Chinese word segmentation, TF‑IDF keyword ranking and lightweight classification to deliver real‑time insights, alerts and actionable reports for thousands of daily user comments.
This article describes how a machine‑learning‑driven monitoring system replaces fixed thresholds with personalized, anomaly‑based detection for business‑level metrics such as network traffic and access volume, detailing the architecture, sample labeling, model training, alarm grading, and operational benefits.
Data scientist Evgeny Bazarov has open‑sourced a 1.589 million‑image NSFW dataset organized into 159 fine‑grained categories, providing GitHub links, download scripts, and a 500 GB storage requirement, enabling researchers to build more precise adult‑content detection models.
This article describes the evolution of 58.com's user‑set service architecture, the transition from MongoDB to RoaringBitmap storage, and the machine‑learning‑driven algorithm optimizations that enable real‑time, multi‑dimensional, and localized push notifications for millions of users.
This article introduces TensorFlow Extended (TFX), illustrating practical TensorFlow examples such as ship trajectory classification, insurance premium adjustments, and car auction pricing, then explains TFX’s data validation, schema generation, model analysis, and deployment options to streamline machine‑learning pipelines.
The article explains content‑based filtering for recommendation systems, covering its basic concepts, feature requirements, implementation using vector representations and cosine similarity, advantages and disadvantages, and supplementary algorithms such as k‑Nearest Neighbor, Rocchio, decision trees, linear classifiers, and Naive Bayes.
This article explains how machine learning, deep learning, and graph‑based techniques are applied to credit anti‑fraud in finance, covering fraud risk characteristics, the anti‑fraud lifecycle, rule limitations, supervised models, common algorithms, neural networks, time‑series models, and graph analytics for detecting individual and group fraud.
This article introduces TensorFlow, explains when neural networks are appropriate, outlines the general workflow for solving image‑based problems, and provides a step‑by‑step Python implementation of a multilayer perceptron that classifies handwritten digits, while also discussing TensorFlow's strengths, limitations, and alternatives.
This article explains how to build a credit scoring card model using logistic regression, detailing the full development workflow—from variable selection and WOE encoding to model training, evaluation metrics like AUC and KS, and deployment considerations for risk management.
This talk by Sogou search advertising researcher Shupeng explains how deep learning techniques are applied to search ad tasks such as automated creative generation and click‑through‑rate prediction, covering system workflow, data pipelines, model evolution from linear models to Wide&Deep and NFM, evaluation metrics, and future directions.
This article explains how automated feature engineering and auto‑modeling techniques dramatically reduce development time and improve performance in fraud‑risk detection, detailing the underlying RFM concepts, feature generation workflow, model selection, evaluation, deployment, and continuous monitoring within a risk‑control platform.
This article presents a comprehensive overview of 58.com’s search efficiency optimization, detailing the business background, ranking framework, data, algorithm, and engineering components, describing the three-stage ranking process, strategy and platform optimizations, feature engineering, model upgrades, and the resulting performance improvements.
This article provides a comprehensive introduction to the Random Forest algorithm, covering its theoretical background, scikit-learn implementation details, practical coding example with the Iris dataset, and a discussion of its advantages, limitations, and typical use cases in machine learning.
Google's 2018 AI review highlights major advances across ethical AI principles, social‑impact projects, assistive technologies, quantum computing, natural‑language models like BERT, perception research, algorithms, TPU hardware, open‑source releases, robotics, healthcare applications, and plans for an even broader impact in 2019.
This article explains a machine‑learning‑driven system that automatically detects and classifies poor‑quality e‑commerce product materials—such as misleading titles, exaggerated benefits, and over‑promotion—to protect consumers, reduce platform risk, and improve conversion rates during major sales events.
This article introduces the Naive Bayes classification algorithm, explains its theoretical basis, demonstrates how to use scikit-learn's GaussianNB class with Python code, evaluates model performance, and discusses advantages, limitations, and practical examples of the method.
This article details Alibaba.com engineers' application of reinforcement learning, attention mechanisms, and weakly supervised techniques to extract product summaries, improve content quality, and significantly raise SEO rankings, supported by offline experiments, online A/B testing, and future research directions.
This article describes the design and implementation of a comprehensive efficiency‑optimization platform at 58.com, detailing its end‑to‑end workflow—from log aggregation and feature extraction through machine learning model training and online experimentation—highlighting modular, configurable, and scalable solutions for multi‑business, multi‑product ranking.
The talk explains how Meituan transformed its O2O hotel search by layering a multi‑stage retrieval pipeline with intent‑aware NLP, then progressively upgrading ranking—from XGBoost to MLPs, feature‑embedding networks, and finally a Wide‑Deep multi‑task model—while tackling data sparsity, diverse scenarios, and deploying the system via TensorFlow‑Serving and the in‑house MLX platform.
The talk by chief architect Luo Hao explains how China Aviation Information tackles heavy legacy systems, non‑standard architectures, and zero‑downtime requirements by using AIOps techniques such as automated configuration discovery, cluster analysis, fault prediction, anomaly detection, event compression and rapid root‑cause automation.
This article introduces core deep‑learning architectures—including MLP, RNN, CNN, auto‑encoders, and RBM—explains common activation and loss functions, and then surveys fourteen influential deep‑learning‑based recommendation algorithms such as FM, wide&deep, deepFM, NCF, GBDT+LR, seq2seq and YouTube DNN, complete with model diagrams and reference links.
This article details the end‑to‑end design, recall and ranking techniques, engineering implementation, and future research directions of Tencent's Yoo video bottom‑page recommendation system, illustrating how large‑scale video recommendation is built from business needs to deep learning models.
This article surveys ten classic recommendation system algorithms—including collaborative filtering, association rules, Bayesian methods, K‑Nearest Neighbors, decision trees, random forests, matrix factorization, neural networks, word2vec, and logistic regression—detailing their principles, mathematical formulas, and practical implementation steps for real‑world applications.
This article presents Ctrip's extensive AIOps exploration, detailing operational challenges caused by massive monitoring data, the evolution of DevOps practices, the design of intelligent anomaly detection and diagnosis systems, practical use cases, and a forward‑looking perspective on the future of AI‑driven operations.
This article walks through the theory of k‑nearest‑neighbors, explains feature selection and normalization, shows how to implement the algorithm with plain JavaScript classes, visualizes results on an HTML canvas, and discusses practical limitations and extensions.
This article presents the design and technical details of 360 Easy Placement, an AI‑driven end‑to‑end advertising platform that automates creative generation, fast review, and optimization, addressing the challenges faced by small‑and‑medium advertisers through data‑rich models, multi‑task learning, and intelligent scene recommendation.
This article presents a comprehensive overview of applying machine learning to recommendation systems, covering fundamental challenges such as user cold‑start, precise interest modeling, collaborative filtering, and both offline and online evaluation methods, while illustrating concepts with numerous diagrams.
In this article, independent Node.js consultant Yoni Goldberg outlines 19 essential skills and topics—from TypeScript and async‑hooks to Kubernetes, blockchain, and machine learning—that developers should explore in 2019 to boost their expertise and stay ahead in the evolving backend ecosystem.
