轮子
平时转型做独立开发与出海调研，最大的痛点就是每天被垃圾信息淹没：刷推特、Product Hunt 和各大社区，充斥着大量套壳工具互吹与营销号焦虑小作文。想找点真正有落地参考价值的技术选型、开源轮子，每天在信息堆里淘金太累。 半个月前基于开源底座二开了一套垂直情报过滤管道，单机部署跑在吃灰的 Linux VPS 上。跑顺了之后每天只推 2~3 条高信噪比卡片，省心很多。但回顾这半个月的二开踩坑，有几个工程层面的细节值得复盘： 1. 语义向量去重的边界与阈值陷阱 最开始想用 pgvector 解决「不同媒体在 48 小时内反复报道同一件事」的冗余。直接用文本做 embedding ，发现余弦相似度阈值如果设在 0.85 ，很多同赛道但不同技术方案的独立项目会被误杀；如果放宽到 0.92 ，换个标题的通稿又漏了进来。
This explanation quotes or summarizes the cited source record. Saving creates a bookmark on this device, with no automatic account sync.
- First observed in our collection
- Oct 25, 2026, 7:30 PM UTC
- Earliest cited publication
- Oct 25, 2026, 2:31 PM UTC
- Last observed
- Oct 25, 2026, 7:30 PM UTC
Evidence in the last complete seven days
- Source mentions
- 0 Previous seven days 1
- Distinct sources
- 0 Previous 1
- Comparable growth
- Not comparable
- Ranking appearances
- 0
Collection observations and undated records are separate from dated publication counts. Observed in 7d：0 · Undated records：0
Collection coverage · Comparable windowsCurrent / previous observed days 7 / 7
Counts use complete UTC days. First observed means first seen by this collection, not when a term was coined.
Current window：2026-10-30 → 2026-11-06 (end excluded) · History 18 days
| Source | Current days | Previous days | Last success |
|---|---|---|---|
| arXiv | 6 | 4 | Nov 6, 2026, 12:30 AM UTC |
| arXiv | 7 | 3 | Nov 6, 2026, 5:15 AM UTC |
| Ask HN | 7 | 4 | Nov 6, 2026, 3:45 PM UTC |
| Ask HN | 7 | 4 | Nov 6, 2026, 3:30 PM UTC |
| Ask HN | 7 | 4 | Nov 6, 2026, 4:00 PM UTC |
| Ask HN | 7 | 4 | Nov 6, 2026, 4:00 PM UTC |
| GitHub | 7 | 4 | Nov 6, 2026, 12:45 PM UTC |
| GitHub Issues | 7 | 4 | Nov 6, 2026, 4:00 PM UTC |
| GitHub Releases | 7 | 2 | Nov 6, 2026, 12:15 PM UTC |
| GitHub | 7 | 4 | Nov 6, 2026, 2:30 PM UTC |
| GitHub Issues | 7 | 4 | Nov 6, 2026, 4:45 PM UTC |
| GitHub Releases | 7 | 4 | Nov 6, 2026, 4:45 PM UTC |
| GitHub | 7 | 4 | Nov 6, 2026, 2:30 PM UTC |
| GitHub Issues | 7 | 4 | Nov 6, 2026, 5:00 PM UTC |
| GitHub Releases | 7 | 4 | Nov 6, 2026, 4:45 PM UTC |
| GitHub Discovery | 1 | 0 | Nov 6, 2026, 4:00 PM UTC |
| GitHub Trending | 7 | 4 | Nov 6, 2026, 1:15 PM UTC |
| 7 | 3 | Nov 6, 2026, 8:30 AM UTC | |
| Hugging Face | 7 | 3 | Nov 6, 2026, 8:30 AM UTC |
| Hugging Face Spaces | 1 | 0 | Nov 6, 2026, 3:30 PM UTC |
| Hacker News | 7 | 4 | Nov 6, 2026, 1:15 PM UTC |
| Hacker News | 7 | 4 | Nov 6, 2026, 4:15 PM UTC |
| Hacker News | 7 | 4 | Nov 6, 2026, 4:15 PM UTC |
| Hacker News | 1 | 0 | Nov 6, 2026, 4:00 PM UTC |
| Show HN | 1 | 0 | Nov 6, 2026, 3:45 PM UTC |
| MIT News | 7 | 3 | Nov 6, 2026, 5:45 PM UTC |
| OpenAI | 7 | 4 | Nov 6, 2026, 11:45 AM UTC |
| Product Hunt | 7 | 4 | Nov 6, 2026, 1:15 PM UTC |
| 量子位 | 6 | 4 | Nov 4, 2026, 10:15 AM UTC |
| OpenRouter | 1 | 0 | Nov 6, 2026, 6:36 PM UTC |
| GitHub Trending | 7 | 7 | Nov 6, 2026, 1:15 PM UTC |
| Hugging Face Models | 1 | 0 | Nov 6, 2026, 6:36 PM UTC |
| HF Daily Papers | 1 | 0 | Nov 6, 2026, 6:36 PM UTC |
| OpenRouter | 1 | 0 | Nov 6, 2026, 6:36 PM UTC |
| Product Hunt | 7 | 7 | Nov 6, 2026, 1:15 PM UTC |
| SEC EDGAR | 7 | 4 | Nov 6, 2026, 10:45 AM UTC |
| Sequoia Capital | 7 | 3 | Nov 6, 2026, 6:45 AM UTC |
| Simon Willison | 1 | 0 | Nov 6, 2026, 3:15 PM UTC |
| TechCrunch | 7 | 4 | Nov 6, 2026, 3:00 PM UTC |
| Contracts Finder | 7 | 4 | Nov 6, 2026, 8:00 AM UTC |
| V2EX | 7 | 4 | Nov 6, 2026, 3:00 PM UTC |
- Counts describe monitored public sources, not the whole web or search demand.
Daily history · up to 90 days
Bars show dated source mentions. The latest UTC day may be incomplete. Zero mentions and missing collection coverage are different conditions.
View accessible daily data
| UTC date | Mentions | Sources | Observations | Ranking appearances |
|---|---|---|---|---|
| 2026-10-23 | 0 | 0 | 0 | 0 |
| 2026-10-24 | 0 | 0 | 0 | 0 |
| 2026-10-25 | 1 | 1 | 1 | 0 |
| 2026-10-26 | 0 | 0 | 0 | 0 |
| 2026-10-27 | 0 | 0 | 0 | 0 |
| 2026-10-28 | 0 | 0 | 0 | 0 |
| 2026-10-29 | 0 | 0 | 0 | 0 |
| 2026-10-30 | 0 | 0 | 0 | 0 |
| 2026-10-31 | 0 | 0 | 0 | 0 |
| 2026-11-01 | 0 | 0 | 0 | 0 |
| 2026-11-02 | 0 | 0 | 0 | 0 |
| 2026-11-03 | 0 | 0 | 0 | 0 |
| 2026-11-04 | 0 | 0 | 0 | 0 |
| 2026-11-05 | 0 | 0 | 0 | 0 |
| 2026-11-06 | 0 | 0 | 0 | 0 |
![Source reference for [程序员] 二开信源管道跑了半个月，聊聊用 pgvector 语义降噪和防重推送踩的几个深坑. Screenshot not yet available.](/media/f490c64e8d8cbbfed34ab183)
Source evidence
Sources in the evidence shown：V2EX 1
V2EX · Published Oct 25, 2026
[程序员] 二开信源管道跑了半个月，聊聊用 pgvector 语义降噪和防重推送踩的几个深坑
平时转型做独立开发与出海调研，最大的痛点就是每天被垃圾信息淹没：刷推特、Product Hunt 和各大社区，充斥着大量套壳工具互吹与营销号焦虑小作文。想找点真正有落地参考价值的技术选型、开源轮子，每天在信息堆里淘金太累。 半个月前基于开源底座二开了一套垂直情报过滤管道，单机部署跑在吃灰的 Linux VPS 上。跑顺了之后每天只推 2~3 条高信噪比卡片，省心很多。但回顾这半个月的二开踩坑，有几个工程层面的细节值得复盘： 1. 语义向量去重的边界与阈值陷阱 最开始想用 pgvector 解决「不同媒体在 48 小时内反复报道同一件事」的冗余。直接用文本做 embedding ，发现余弦相似度阈值如果设在 0.85 ，很多同赛道但不同技术方案的独立项目会被误杀；如果放宽到 0.92 ，换个标题的通稿又漏了进来。
View source record →
Review priority 10 / 100
This score orders research candidates. It is not search volume, a forecast or verified commercial demand.
Evidence
0 / 30Distinct dated articles or first observations; repeat crawling adds nothing.
Publisher diversity
0 / 25Conservative owner grouping; syndicated copies count once.
Comparable change
0 / 20Comparable complete weeks only; no percentage from a zero baseline.
Recent evidence
2.2 / 15Latest publication or first observation, not repeated fetch time.
Name specificity
8 / 10Explicit entities and technical concepts outrank ambiguous acronyms.
Limitations and interpretation
- First observation is a collection milestone, not proof that a word or product is new to the world.
- Counts describe monitored public sources, not the whole web or search demand.
- First observed here is not the invention date of a name or idea.
- Candidates come from public titles, summaries and entity URLs, not verified full-text analysis.
- Owner grouping is conservative; undisclosed syndication may remain.
- The previous week is too small for a reliable percentage change.
- Current evidence comes from one publisher or owner.