Author

Ma Wei Says

Follow me! Discussing software architecture and development, AIGC and AI Agents... Sometimes sharing insights on IT professionals' life experiences.

Articles

Likes

208

Views

Comments

Latest from Ma Wei Says

31 recent articles

Ma Wei Says

Feb 23, 2025 · Artificial Intelligence

How Microsoft’s PIKE‑RAG Builds Knowledge‑Driven AI Across Four Stages

The article explains Microsoft’s open‑source PIKE‑RAG system, detailing its four progressive stages—from knowledge‑base construction to creative multi‑agent reasoning—while describing the underlying modules, chunking strategies, multi‑granularity retrieval, and code snippets that enable specialized domain understanding and inference.

AI RetrievalKnowledge GraphLLM

0 likes · 11 min read

How Microsoft’s PIKE‑RAG Builds Knowledge‑Driven AI Across Four Stages

Ma Wei Says

Feb 22, 2025 · Artificial Intelligence

Pick the Best Vector Similarity Strategy in LangChain: Euclidean, Inner Product, Cosine

This guide explains how to configure and import the appropriate DistanceStrategy in LangChain, compares Euclidean distance, maximum inner product, and cosine similarity, and outlines their formulas, advantages, and typical use‑cases for vector‑based retrieval.

AI embeddingsLangChainRAG

0 likes · 6 min read

Pick the Best Vector Similarity Strategy in LangChain: Euclidean, Inner Product, Cosine

Ma Wei Says

Feb 21, 2025 · Artificial Intelligence

How PIKE‑RAG Boosts Retrieval‑Augmented Generation for Industrial AI

PIKE‑RAG, a Retrieval‑Augmented Generation framework from Microsoft Research, tackles knowledge source diversity, one‑size‑fits‑all limitations, and LLMs' lack of domain expertise by building multi‑layer heterogeneous graphs, task‑driven modular pipelines, and a staged L0‑L4 system for more accurate industrial AI responses.

AIKnowledgeGraphLLM

0 likes · 5 min read

How PIKE‑RAG Boosts Retrieval‑Augmented Generation for Industrial AI

Ma Wei Says

Feb 19, 2025 · Fundamentals

Safe List Operations: Remove, SubList Casting, and Efficient Traversal in Java

This article explains why modifying a List inside a foreach loop causes ConcurrentModificationException, shows the proper ways to remove elements using iterators or lambda expressions, warns against casting subList to ArrayList, clarifies correct toArray and Arrays.asList usage, compares LinkedList and ArrayList performance, and recommends the most efficient traversal techniques.

IteratorJavaList

0 likes · 9 min read

Safe List Operations: Remove, SubList Casting, and Efficient Traversal in Java

Ma Wei Says

Feb 18, 2025 · Databases

Which Data Types Should You Use for Money in Java & MySQL? Best Practices

This guide explains why VARCHAR, float, and double are unsuitable for monetary values, and recommends using BIGINT/Long for cent‑level precision or BigDecimal/DECIMAL for higher precision, with practical tips and code examples for Java and MySQL.

BigDecimalJavaMonetary

0 likes · 6 min read

Which Data Types Should You Use for Money in Java & MySQL? Best Practices

Ma Wei Says

Feb 14, 2025 · Databases

Master MySQL Regex: All REGEXP Functions Explained

This guide explains MySQL's regular‑expression capabilities—including REGEXP, NOT REGEXP, REGEXP_LIKE, REGEXP_REPLACE, REGEXP_SUBSTR, REGEXP_INSTR and their RLIKE equivalents—covering syntax, multibyte safety, match‑type flags, and practical usage examples.

MySQLSQLregex

0 likes · 7 min read

Master MySQL Regex: All REGEXP Functions Explained

Ma Wei Says

Feb 13, 2025 · Artificial Intelligence

Master AI Prompting: 5 Proven Techniques to Unlock Accurate Outputs

This guide presents five practical prompting techniques—including structured output, role‑playing, visual conversion, multi‑turn refinement, and multilingual handling—plus industry‑specific examples and common pitfalls, helping users craft precise commands for AI models like DeepSeek.

AI promptinglarge language modelsprompt engineering

0 likes · 8 min read

Master AI Prompting: 5 Proven Techniques to Unlock Accurate Outputs

Ma Wei Says

Feb 12, 2025 · Databases

Understanding Redis LRU and LFU: How Memory Eviction Works

This article explains Redis's LRU and LFU eviction algorithms, their implementation details, code structures, and practical guidance on choosing the right policy based on workload characteristics.

LFULRURedis

0 likes · 6 min read

Understanding Redis LRU and LFU: How Memory Eviction Works

Ma Wei Says

Feb 11, 2025 · Databases

Master Redis Memory Eviction: All 8 Policies and Configuration Tips

This article explains Redis's eight memory eviction policies, their ideal use cases, and provides step‑by‑step commands and configuration examples for querying and setting both the eviction strategy and maximum memory limits.

CacheDatabase ConfigurationLFU

0 likes · 6 min read

Master Redis Memory Eviction: All 8 Policies and Configuration Tips

Ma Wei Says

Feb 10, 2025 · Databases

How Redis Handles Expired Keys: Periodic vs Lazy Deletion Strategies

This article explains Redis's two expiration mechanisms—periodic scanning with configurable frequency and lazy deletion on client access—detailing their configurations, execution steps, performance trade‑offs, and replication pitfalls to help developers manage memory efficiently.

Lazy DeletionMemory ManagementRedis

0 likes · 5 min read

How Redis Handles Expired Keys: Periodic vs Lazy Deletion Strategies