This tutorial explains how to transform raw text into tensor representations using one‑hot encoding, Word2Vec, and FastText, provides step‑by‑step code examples, discusses their advantages and drawbacks, and shows how to visualize embeddings with TensorBoard.
Before training large language models like LLaMA, MetaAI applies a multi‑stage CCNet pipeline that crawls web data, stores it in WET format, deduplicates paragraphs, detects and filters languages using fastText, and further refines content by similarity to Wikipedia and citation‑based linear models.
This article shows how the compress‑fasttext Python library can shrink a 7 GB fastText word‑embedding model to about 21 MB—a 300‑fold reduction—while preserving almost the same accuracy on downstream NLP tasks, and explains the underlying compression techniques, usage examples, and evaluation results.
This article explains practical techniques for improving NLP model accuracy on massive corpora, covering challenges of multi‑field text, word‑embedding choices, a fasttext‑based regression demo with book‑review data, feature engineering tricks, and a comparison with tf‑idf + LASSO.
This article describes how 58.com’s AI Lab built and continuously improved the QABot intelligent customer‑service system by designing classification and retrieval models, evaluating FastText, LSTM‑DSSM, BERT and a self‑developed SPTM framework, and finally fusing them to boost answer rates and user experience.
This article details the BangBang intelligent customer service system's overall architecture, core capabilities, knowledge‑base construction, and successive model upgrades—from FastText to TextCNN, Bi‑LSTM, and model fusion—showing how each iteration improved accuracy, recall, and F1 scores toward a stable 95% performance level.
This article describes how a large‑scale international hotel platform reduced room‑type merging errors and user complaints by applying rule‑based methods, text‑similarity algorithms (Jaccard, LCS, N‑Gram) and supervised machine‑learning classifiers such as fastText to standardize and merge heterogeneous room‑type data.
This article reviews the evolution of word and sentence embeddings, covering foundational theories like vector semantics and distributional hypothesis, practical models such as Word2Vec, GloVe, fastText, Skip‑Thought, Quick‑Thought, and evaluation techniques, while offering implementation tips and real‑world use cases.
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.
The talk outlines how Beike built a real‑estate conversational AI platform, covering the market need for dialogue systems, the five technical challenges, data‑driven intent and slot extraction, model choices such as FastText and Bi‑LSTM‑CRF, a three‑layer system architecture, multi‑intent handling, and future directions like 4D viewing and an internal AI dialogue platform.
