This article explores the evolution and monetization of operations data, outlines a four‑stage management process—from data discovery to modeling, ingestion, and monetization—highlights key scenarios such as intelligent monitoring and root‑cause analysis, and offers practical recommendations for building an effective ops data platform.
This case study details how 58.com responded to the COVID‑19 pandemic by conducting market and user research, analyzing business opportunities, and creating a comprehensive design system—including brand, visual, illustration, and AI‑driven tools—to enhance service breadth, improve user experience, and drive growth across online and offline channels.
This presentation details ZTO Express’s rapid rise from a modest startup to the world’s largest courier by exploring its technology‑driven business model, crowd‑funded expansion, electronic waybills, smart routing, AI customer service, employee equity schemes, and future digital logistics strategies.
This article provides a detailed introduction to knowledge graphs, covering their definition, why they are needed, the four basic triple types, construction pipelines (including data sources, crowdsourced vs automated methods, and schema versus data layers), storage and query techniques using graph and relational databases, and their practical applications such as enhancing precision, diversity, and explainability in recommendation systems through models like DKN, RippleNet, and graph neural networks.
This article details Likee's short‑video recommendation pipeline, covering the evolution of its feature‑engineering framework, the use of Flink for minute‑level statistical and second‑level session features, the integration of automatic statistical features into DNN models, multimodal feature extraction, and the cache‑tensor technique that dramatically improves online inference performance.
Tencent joins the Linux Foundation Public Health initiative, open‑sourcing tools like the international COVID‑19 health module, an AI self‑triage assistant, and a severe‑case prediction model with Zhong Nanshan’s team, empowering developers, hospitals, and public‑health agencies worldwide to combat the pandemic.
The announcement introduces an AI‑driven intelligent customer service competition, highlighting the importance of text matching and classification in NLP, describing 58.com’s existing system, providing a real‑world dataset, and inviting participants to develop precise models using the latest deep‑learning techniques.
This newsletter blends deep dives into Vue 3 RFC processes, cloud‑gaming opportunities, robust error‑code design, AI‑driven SQLFlow and Pipcook tools, plus handy regex and FIGlet utilities, offering developers practical knowledge across front‑end, back‑end, and emerging technologies.
This case study analyzes the Bank 4.0 transformation of traditional Chinese banks, detailing industry characteristics, historical challenges, open‑banking drivers, the ABCDII technology framework, DevOps tooling, metric systems, and a future ecosystem vision to guide digital and operational improvement.
This article examines the policy backdrop, industry challenges, and JD Health’s AI‑driven smart triage solution—including its knowledge‑base, NLP, and multimodal fusion algorithms—and showcases its impact during the COVID‑19 surge while outlining future platform and lifecycle extensions.
This guide walks through creating a browser‑based face login system by capturing video with getUserMedia, processing frames with Baidu's facial analysis API and the Python face_recognition library, and integrating the workflow with front‑end HTML/JavaScript and back‑end handling of image data.
The article examines the hidden costs of traditional Selenium scripting, explains how no‑code Selenium platforms simplify test creation, improve maintainability with AI‑driven self‑healing, streamline execution through built‑in CI/CD integration, and outlines scenarios where native Selenium remains preferable.
The article explains how to reduce the impact of frequent UI changes on automated testing by using visual testing techniques, introduces the Watcher platform’s SDK, image‑comparison services, real‑time comparison, and various algorithms—including pixel, template, content, and AI‑based matching—to improve test stability and efficiency.
In this interview, DYG team member Wang He introduces his teammates, explains why they formed the team, and shares detailed model‑fusion techniques—including input variation, diverse model architectures, and training‑target differences—to boost competition scores during the final stage.
The AMAP-TECH algorithm competition invites participants to develop AI models that analyze in-vehicle video sequences to determine dynamic road conditions, offering detailed dataset specifications, evaluation metrics, expert judges, schedule, and prize information for researchers in computer vision and traffic analytics.
The article provides a comprehensive overview of FPGA technology, detailing its architecture, advantages over ASICs, current global market size, growth drivers such as 5G and AI, and the challenges and opportunities faced by domestic Chinese manufacturers in a market dominated by Xilinx and Intel.
This article surveys the role of social tagging (folksonomy) in modern recommendation systems, describing how user‑generated tags form a three‑dimensional "tag cube" that can be combined with rating matrices, and reviewing a range of algorithms—including neighbor‑based, ranking (FolkRank/SocialRank), content‑based, linear regression, and matrix‑factorization approaches—while also discussing tag selection, noise handling, and scalability challenges.
Tencent’s AI White Paper outlines the concept of ‘ubiquitous intelligence’, analyzing macro trends, policy support, supply‑demand dynamics, core technologies such as computer vision and NLP, and the pandemic‑driven acceleration of AI applications across healthcare, smart cities, manufacturing, and digital content.
The article examines how AI-driven mosaic removal tools, leveraging GAN technology, are reviving uncensored adult video content, sparking privacy, copyright, and ethical debates while detailing the technical challenges, industry reactions, and the rise of related software among Chinese users.
This technical talk introduces video object segmentation tasks, reviews leading datasets and state-of-the-art deep learning models, and shares practical network design rules and performance‑boosting techniques, presented by Prof. Wang Xinggang as part of Alibaba's MEDIA AI challenge series.
An advanced multi‑level, multi‑modal search and graph engine for Youku processes text, voice, image and video queries across hierarchical video elements, using combined vector and inverted indexes to merge cross‑level and cross‑modal results, while a distributed knowledge‑graph layer enables multimodal graph traversal for billion‑scale video retrieval.
Alibaba Amap’s AMAP‑TECH competition invites participants to develop AI computer‑vision models that classify real‑time road conditions—smooth, slow, or congested—from short sequences of dash‑cam images, using a labeled dataset of 1,500 training sequences and a weighted F1‑score evaluation, with cash prizes up to ¥60,000.
This article explains the fundamentals of recommendation systems, discusses early hotness‑based approaches, describes modern architectures with recall and ranking stages, reviews collaborative‑filtering techniques, matrix factorization, deep learning models such as NCF and NeuMF, and details how these methods are applied and optimized for live‑streaming services.
AWS CodeGuru, an AI‑driven code review service, automatically analyzes code in GitHub or Bitbucket, offers performance and error suggestions, and helps large teams maintain high‑quality software without the overhead of manual reviews.
The article presents a detailed technical overview of Youku's video search system, covering multi‑modal inputs, multi‑level element indexing, face search, cross‑level and cross‑modal retrieval, and the design and applications of a multimodal graph engine with knowledge‑graph integration.
This guide shows pandemic graduates how to create AI‑generated personalized graduation photos with Huawei ModelArts, locate optimal rental housing using Python‑driven web scraping and Gaode maps, and accelerate job preparation through Huawei Cloud Classroom courses and coding practice, turning a challenging season into a tech‑enabled opportunity.
This guide walks researchers through selecting a Tencent Cloud GN7 GPU instance, installing NVIDIA drivers, CUDA 10.2, cuDNN, setting up PyTorch and Jupyter, and configuring remote development with PyCharm, enabling efficient, cost‑effective AI development on a Tesla T4 GPU server.
This article explores how artificial intelligence is reshaping test automation by addressing maintenance bottlenecks, enhancing exploratory testing, and offering practical guidance for teams to adopt AI‑driven tools while remaining accountable for accuracy and results.
This article shares the Xiaomi AI Lab NLP team's experience in the NLPCC lightweight language model competition, discusses efficiency challenges of large pretrained models like BERT, and details practical inference optimizations—including model distillation, batching, FP16 quantization, and FasterTransformer integration—that dramatically reduce latency and hardware costs in production.
This article examines the current status, strategic importance, and underlying reasons for federated learning and blockchain, compares their similarities and differences, and explores how their complementary strengths can be combined to create trusted, privacy‑preserving, and value‑transfer solutions in the digital economy.
This article explores Alibaba’s effort to automate code generation by converting product requirements into standardized design language and leveraging AI and machine learning to bridge the gap between natural language specifications and executable frontend code.
This tutorial walks you through using Pipcook’s text‑creation pipeline to train a TensorFlow.js LSTM model that generates Chinese poems from a given seed, covering plugin configuration, GPU training, model evaluation, and Node.js inference.
Three tech veterans—software engineer Qingrun, quantum researcher Hao Shaogang, and investor‑entrepreneur Li Zhuohuan—share their career paths and insights on AI, cloud, 5G, edge and quantum computing, emphasizing curiosity‑driven learning, real‑world problem solving, and interdisciplinary innovation.
Pipcook 1.0, an open‑source machine‑learning toolkit designed for front‑end developers, brings seamless Python integration, a flexible plugin system, and easy‑to‑use pipelines and PipApp experiments, enabling JavaScript engineers to build, train, and deploy AI models directly in their web projects.
This article explores the rapid rise of AI in the frontend ecosystem, compares native and web‑based inference solutions, evaluates frameworks such as TensorFlow JS, ONNX JS, WebNN and the new MNN.js, and shares performance data, code examples, and future directions for cross‑platform AI deployment.
The article analyzes the rapid shift from "Made in China" to "Intelligent Manufacturing" driven by industrial internet, 5G, AI and big data, highlighting policy evolution, case studies across light industry, liquor production and hazardous chemicals, and Tencent Cloud's strategic role in enabling digital transformation.
Meituan uses AI to solve the massive combinatorial routing problem of its food‑delivery network, optimizing rider paths, aiding merchants, providing voice‑controlled ordering for the visually impaired, and deploying computer‑vision, robotics, autonomous vehicles and drones to streamline logistics while keeping riders focused on short‑distance deliveries.
The article reviews the open‑source qa_match tool from 58.com, detailing its deep‑learning based question‑answer matching architecture, hierarchical knowledge‑base support, lightweight pre‑training model SPTM, and practical applications, while summarizing the live‑stream presentation and Q&A session.
The article surveys the rapid growth of AI, presenting adoption statistics, a Forrester‑predicted market surge, a ranked list of the ten most popular AI technologies with vendor examples, and the key obstacles enterprises face when implementing AI solutions.
The four‑day online 2020 Beijing Zhiyuan AI Conference (June 21‑24) convened five Turing Award laureates, over 100 AI experts, and 19 specialized forums to explore causal inference, deep learning, ethics, healthcare, transportation, entrepreneurship, and a broad range of technical topics, with plenary discussions on cause‑effect science, industry trends, and AI safety.
This article describes the motivation, advantages, challenges, overall architecture, and natural‑language‑understanding algorithms of Ctrip’s one‑stop chatbot that consolidates multi‑service line support into a single entry, enabling seamless travel‑related queries, recommendations, and bookings.
This article presents a comprehensive overview of 58's recruitment recommendation system, detailing business challenges, multi‑stage recall strategies, vector‑based deep retrieval, cost‑sensitive loss design, session optimization, online incremental training, extensive offline and online evaluations, and practical lessons for future improvements.
Tencent Cloud’s senior engineer Li Kaibin outlines the cloud‑based face‑effects platform, detailing its AI‑driven features such as face fusion, beauty, virtual makeup, segmentation and age‑gender transformation, the CNN‑based model training pipeline, a layered service architecture with elastic scaling and robust monitoring, and future expansions into video effects, international regions and low‑code integration.
The article examines the rise of AI‑generated deepfake media, its legal treatment under China's new Civil Code and other regulations, the technical ease of creating such content, and the combined efforts of detection technology, ethical education, and legislation to curb its harmful impact.
The article introduces qa_match V1.1, an open‑source deep‑learning QA matching tool that adds one‑level knowledge‑base support, releases a lightweight Bi‑LSTM pre‑trained language model (SPTM), and provides detailed architecture, training data, performance benchmarks, future plans, and contribution guidelines.
The ICRA 2020 TrajNet++ competition challenged teams to predict 4.8‑second pedestrian paths from 3.6‑second observations, and Meituan’s winning solution used a Seq2Seq world‑model that encodes past trajectories, updates a spatio‑temporal interaction map, and decodes future positions, achieving a 1.24 m final displacement error and demonstrating readiness for real‑world unmanned delivery.
The article explains how advertising inventory forecasting—predicting how many users will see a specific ad—poses challenges due to fluctuating traffic and user segmentation, and describes a new deep spatial‑temporal tensor factorization model that dramatically improves prediction accuracy, scalability, and robustness for large‑scale ad platforms.
Alibaba is advancing cargo‑focused autonomous driving by building a fine‑grained scenario library and the AutoDrive platform to automate perception, planning, and control algorithms, targeting low‑speed L4 logistics robots while navigating sensor, hardware, and regulatory challenges and anticipating broader L2 adoption and tighter hardware‑software co‑design.
The article introduces qa_match V1.1, an open‑source, Apache‑licensed lightweight question‑answer matching system that adds a simple pre‑trained language model (SPTM), supports one‑level knowledge bases, details model architecture, training resources, performance benchmarks, future plans, and contribution guidelines.
This article introduces the challenges of traditional centralized recommendation systems, explains the principles and implementations of federated recommendation algorithms—including vertical and horizontal federated matrix factorization and factorization machines—using WeBank’s open-source FATE platform, and discusses cloud services, practical use cases, and performance benefits.
Feature engineering, which occupies the majority of data scientists' time, is essential for building high‑performing machine‑learning models and involves careful data quality control, diverse construction techniques, rigorous selection, and emerging automation efforts, all of which demand domain expertise and systematic practice.
This article explains how to use deep‑learning object detection to automatically recognize UI components in design drafts, generate a smart JSON description, and convert it into component‑based front‑end code, covering problem analysis, dataset preparation, algorithm selection, model training, evaluation, and deployment.
The article refutes false claims that WeChat monitors chats, explains how SpaceX’s Crew Dragon and Falcon 9 rely on custom Linux systems for flight software, and introduces Baidu’s open‑source quantum machine‑learning framework Paddle Quantum built on the PaddlePaddle AI platform.
This article presents a comprehensive overview of more than forty Baidu open‑source projects, ranging from JavaScript charting libraries and UI frameworks to distributed configuration platforms, mobile testing tools, RPC frameworks, big‑data visualization, and AI software, highlighting their main features and typical use cases.
This article compiles over 70 open‑source projects released by Baidu, spanning front‑end libraries, backend frameworks, mobile tools, data‑visualization, distributed systems, AI platforms, and more, offering brief descriptions to help developers discover which of these resources they have already used.
The article explains how a well‑defined technology roadmap, encompassing architecture, data governance, storage, analytics, and automation, is essential for aligning tools and techniques with business goals to achieve a successful, outcome‑driven enterprise data strategy.
The article recaps Alibaba Cloud’s five‑day Vision AI training camp, detailing the core technologies, two hands‑on projects (a smart photo album and an ID‑card OCR system), student showcase code snippets, personal reflections, and the announcement of the next session.
The WISE x City New Engine summit gathered industry leaders to explore how AI, industrial internet, 5G and digital transformation are driving China’s new‑infrastructure strategy, revealing opportunities, challenges, and practical insights for retailers, healthcare, and smart‑city development.
The article examines how China’s “new infrastructure” drive, combined with government‑enterprise collaboration and AI‑powered big‑data solutions, is transforming early‑stage smart communities into integrated, one‑hour life circles that address residents’ evolving service needs.
This article explains how enterprises can classify structured, semi‑structured and unstructured data as “dark data”, why traditional RPA struggles with it, and how AI technologies like NLP, computer vision and machine learning—exemplified by Automation Anywhere’s IQ‑Bot—enable end‑to‑end automation of hidden information.
This essay traces humanity's technological journey—from early agricultural and industrial revolutions through the rise of computing, cloud, and artificial intelligence—highlighting how demand, generalization, cost reduction, and safety shape each breakthrough and drive societal transformation.
When web applications become highly dynamic, traditional Selenium locators become fragile and costly to maintain, but leveraging AI and machine‑learning‑based object‑recognition frameworks can dramatically reduce maintenance overhead and accelerate test creation by intelligently adapting to UI changes.
WeChat’s Scan‑to‑Identify system now trains its CNN models across multiple GPUs using Horovod’s data‑parallel, synchronous Ring All‑Reduce architecture built on MPI and NCCL, cutting training time from several days to under one day while maintaining accuracy, and future work will target I/O and further scaling.
The AIS 2020 Conference, co‑hosted by the Beijing Academy of Artificial Intelligence and Meituan, showcased 74 top ACL, IJCAI and SIGIR papers across 15 sessions on NLP, AI and IR topics, streamed free online on May 23‑24 2020 with keynote speakers from leading Chinese universities.
TAL Education's AI Engineering Institute recently had seven of its machine‑learning papers selected for prestigious conferences such as AIED 2020, EDM 2020, ICASSP 2020 and WWW 2020, showcasing advances in speech recognition, data mining, multimodal learning and educational AI applications.
The article outlines five major DevOps trends for 2020—including AI‑driven automation, the rise of Golang, serverless architecture, cloud‑native practices, and the growing focus on DevSecOps—explaining how each will reshape software delivery and operational efficiency.
The article examines how smart community initiatives, driven by AI, big data and cloud technologies, can overcome regional imbalances and information silos to create integrated, efficient, and resident‑friendly services that form a one‑hour lifestyle ecosystem.
The May 2024 Huawei Cloud Community newsletter curates expert articles and tutorials on AI dark data handling, Apache CarbonData advances, agile retrospectives, IoT Greengrass on Raspberry Pi, Kubernetes networking, WAF anti‑scraping, openEuler contributions, Kunpeng cloud case studies, and ModelArts practical projects.
This article presents a study on enhancing text representation and clustering in a small‑sample setting for 58's second‑hand car intelligent customer service by introducing a Bi‑LSTM based pre‑training language model and an improved Deep Embedded Clustering (DEC) algorithm, demonstrating significant gains in accuracy, silhouette score, and answer‑rate through extensive experiments.
This article describes the design, model selection, experimental evaluation, and engineering optimization of an AI‑driven post‑sale question‑answering system for hotel services, covering FAQ construction, intent detection, deep‑learning matching models such as DSSM, ESIM, BERT, and their performance and latency trade‑offs.
This article reviews the Text-to-SQL semantic parsing task, covering its motivation, dataset landscape, major model architectures such as pointer networks, sequence‑to‑set, and grammar‑based approaches, evaluation metrics, the newly built DuSQL dataset and DuParser system, real‑world deployments, and remaining research challenges.
This article explores how large‑scale trajectory mining can overcome the timeliness limits of traditional street‑sweeping data collection, detailing the underlying principles, technical challenges such as vehicle‑type detection and map‑matching, and practical solutions ranging from rule‑based filters to advanced AI models.
Ren Haibing, Alibaba’s senior algorithm expert, explains how deep‑learning AI matting—combining salient object detection, semantic and instance segmentation in a two‑stage hard‑then‑soft pipeline—replaces traditional methods, achieves hair‑level detail and temporal consistency, scores 84.3% on Cityscapes, powers large‑scale video‑person extraction, and relies on human review to meet growing industry demand.
A Chinese netizen used AI to colorize, upscale, and increase the frame rate of 1920 Beijing footage, sparking massive online interest, while the article explains the underlying AI interpolation methods, Gigapixel AI's role, and the technical challenges of restoring century‑old video.
The article details how a data analyst scraped comments from Zhihu, Weibo, and Bilibili about the 2020 Youth Day video “后浪”, applied Baidu AI sentiment analysis and Python word‑cloud generation, and compared the differing emotional tones across platforms.
This article reviews how large‑scale pretrained language models have reshaped NLP, outlines the challenges of applying them to machine translation, introduces the APT framework and the GRET architecture for better encoder‑decoder integration, and reports experimental gains and future research directions.
This article reviews the major challenges of AI—data, model, and knowledge—explains why automated machine learning and neural architecture search are crucial, analyzes weight‑sharing NAS algorithms and their instability, presents various improved DARTS‑based methods, and discusses experimental results and future research directions.
Carrefour’s new AI‑driven fast‑pick warehouse system, built by Suning, cuts picking time by half and enables over 95% of more than 1,000 daily orders to be delivered within an hour, illustrating the power of big‑data logistics in modern retail.
This article presents a comprehensive overview of graph‑based recommendation techniques, including collaborative filtering, graph embedding, side‑information enhanced embeddings, two‑tower DSSM models, and practical cold‑start strategies from Alibaba and Airbnb, followed by a mixed model and Q&A session.
Ctrip’s flight search service processes two billion daily queries by employing a multi‑level Redis cache, machine‑learning‑driven TTLs, distributed pooling and overload protection, AI‑based anti‑scraping, and robust load‑balancing across three data centers, delivering sub‑second latency, up to three‑fold throughput gains and significant cost reductions.
This article examines the frequent pitfalls of recommendation systems—misleading metrics, over‑optimizing precision, data leakage, feature inconsistencies, and distribution bias—that cause offline AUC improvements to translate into lower online CTR and CPM, and offers practical mitigation strategies.
This article explains the classification, value, and detailed interpretation of the first national shared intelligence standard, describes its technical background and solutions such as Trusted Execution Environments and Secure Multi‑Party Computation, and showcases real‑world applications and industry recognitions.
In a live Q&A, Alibaba Amap’s chief scientist Ren Xiaofeng explained how computer‑vision algorithms underpin high‑precision map creation, AR navigation, visual localization and sensor fusion, discussed current hardware limits, deep‑learning bottlenecks, 5G’s role, edge‑cloud cooperation, and offered career advice for transitioning researchers.
This article examines the limitations of traditional new media marketing, explains reinforcement learning fundamentals, and presents a six‑step technical solution—including problem modeling, algorithm selection, action, state, reward design, and model training—that uses RL to optimize budget allocation and achieve over 35% improvement in campaign effectiveness while reducing costs.
The article examines China's push to develop data factor markets and showcases how Suning’s integration of online and offline data, big‑data analytics, and AI is revitalizing traditional retail, illustrating the broader impact of digital transformation on the post‑pandemic economy.
A data product succeeds by aligning three core layers—an application layer that defines the business goal, a data layer that gathers and organizes high‑quality datasets, and an algorithm layer that applies statistical and AI techniques—to transform raw data into actionable solutions for users, enterprises, or governments.
This tutorial walks you through creating a desktop Python chatbot using the Turing API, covering required libraries, authentication, code implementation for both basic and Selenium‑enhanced versions, and packaging the final program into an executable.
This weekly tech roundup reviews the past ten years of frontend evolution, forecasts the next decade, highlights Vue 3’s beta release and ES2020 finalization, and explores practical CSS math functions and Webpack’s Module Federation, offering concrete insights for developers aiming to stay ahead.
This article explains how Pipcook leverages TensorFlow.js to create a JavaScript‑friendly machine‑learning pipeline for front‑end engineers, addressing skill gaps, data handling, model training, deployment options, and future roadmap to accelerate intelligent front‑end development.
This article surveys the latest Java SE open‑source landscape, compares JDK options, highlights Alibaba Dragonwell’s innovations, and explores emerging OpenJDK trends such as JFR, ZGC, Graal, Loom, as well as Java’s evolution for cloud‑native, AI, and polyglot programming.
This article examines DeepL’s translation system, comparing its neural‑network‑driven output to Google and other services, detailing its Icelandic HPC infrastructure, data collection, architectural choices, language support, strengths, limitations, and expert opinions on why it often delivers more natural translations.
This article presents a comprehensive overview of Youku's video search system, covering business background, evaluation metrics, system and algorithm frameworks, relevance and ranking feature engineering, dataset construction, semantic matching, multimodal video understanding, and practical case studies that illustrate the impact of deep learning and AI techniques on search performance.
This article outlines an AI-powered workflow for constructing and enriching a business knowledge base in intelligent customer service, covering preprocessing, intent detection, deep and shallow semantic feature engineering, hierarchical bucket clustering, and automated summary extraction to improve FAQ coverage and reduce manual workload.
This article explains how multi‑armed bandit methods such as Upper Confidence Bound, Thompson Sampling, and their contextual extensions can address cold‑start, diversity, and bias problems in large‑scale recommendation systems, describing practical update mechanisms, offline evaluation techniques, and deployment experiences at Ctrip.
This article details a step‑by‑step tutorial for creating a celebrity face‑matching application that crawls star information, stores images and metadata in a MySQL database, and uses Baidu's facial recognition API to compare uploaded photos, outputting similarity scores and matched celebrity details.
The talk presents Alibaba’s hierarchical video‑category system and a multimodal classification pipeline—leveraging EfficientNet, NeXtVLAD fusion, attention‑dropping augmentation, and MoCo contrastive learning—that together boost cold‑start recall by 43%, improve program classification over 20%, and set the stage for larger models and advanced unsupervised methods.
This article details the evolution of Feizhu's "Guess You Like" ranking system, moving from a linear FTRL model to several end‑to‑end deep learning versions—including PALM, FB‑PALM, and GLA—highlighting technical challenges, architectural changes, and measurable performance gains.
Edge‑cloud collaboration amplifies the value of edge and cloud computing by integrating distributed resources, data processing, AI inference, and network slicing across diverse scenarios such as industrial IoT, smart homes, and MEC, enabling low‑latency, high‑bandwidth services, optimized resource management, and new business models.
This article traces a decade of frontend evolution—from early visual tools and module‑based workflows to modern AI‑driven automation—explaining why developers must shift toward intelligent, low‑code solutions to keep pace with exploding business demands and ever‑changing devices.
This article analyzes Huajiao Live's end‑to‑end recommendation pipeline, covering basic concepts, recall and ranking algorithms—including collaborative filtering, matrix factorization, deep learning models—and multi‑objective optimization, while detailing the engineering workflow for training, deployment, and real‑time serving in a live‑streaming environment.
