Unverified50% confidenceFactExact time
在 Lucene 和 Elasticsearch 中，词典实际使用 FST（有限状态转换器），它是 DAWG 的推广形式，不仅能判断词是否存在还能关联输出值
1
Sources
50%
Confidence
Long-term
Relevance
8/11/2026
First Seen
Sources
Related Claims
UnverifiedFST (Finite State Transducer) achieves extremely high compression ratios by sharing prefixes and suffixes, making it particularly suitable for handling collections of strings with many common prefixes.65% similarUnverified主流Embedding模型（如BGE、text-embedding系列）通常基于BERT架构的编码器（Encoder-only），通过对比学习让语义相近的句子在向量空间中靠拢；而GPT类生成模型基于解码器（Decoder-only）59% similarUnverifiedTools such as Semantic and Difftastic are exploring the direction of AST-based semantic diff.58% similarUnverifiedLyra is a meta-prompt used to generate high-quality prompts, rather than a software tool.56% similarUnverifiedTypesense用C++编写，强调易用性；Tantivy是Rust编写的全文搜索库，相当于Apache Lucene的Rust版本55% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/731175API
curl https://kongchang.com/api/v1/knowledge/claims/731175MCP
get_claim(id=731175)