2.3 版本中的搜索改进

新闻和活动 公告
1

对于即将发布的 2.3 版本，我们添加了一项新功能，并对帖子搜索索引方式进行了多项修复，这将使搜索结果更加优化。

1. 分类的搜索优先级

Screenshot%20from%202019-04-04%2011-20-00
Screenshot%20from%202019-04-04%2011-20-00631×646 47 KB

每个分类都可以配置搜索优先级，您可以在创建或编辑分类时的 Settings（设置）选项卡中找到该选项。新增了五个优先级等级：ignore（忽略）、very low（非常低）、low（低）、high（高）和 very high（非常高）。这些等级通过将预配置的权重乘到每个结果的搜索排名中来生效，并可通过控制台中的隐藏站点设置进行配置。

例如，将分类的搜索优先级配置为 very_high 会使其搜索排名提升 40%，而配置为 very_low 则会使其搜索排名降低 40%。将分类的搜索优先级设置为 ignore 会将其从搜索结果中移除。不过，您仍然可以通过高级搜索将搜索范围限定在该分类内，从而搜索其中的帖子。请注意，搜索优先级不会继承，这意味着即使父分类被配置为忽略，其子分类仍然可被搜索。

2. 搜索结果的改进与修复

  1. 搜索相关性已更新，在排名时会考虑文档长度。此前，搜索排名仅根据与给定搜索词匹配的次数最多来决定。这对搜索结果来说存在问题，因为我们发现较长的帖子更有可能因为匹配次数较多而排名靠前。因此，我们已改为在排名搜索结果时考虑文档长度。
    FIX: Relevance search will now consider document length in ranking. · discourse/discourse@e87ca59 · GitHub

  2. 提高了用于生成搜索索引的原始数据质量。PERF: Improve quality of `PostSearchData#raw_data`. (#7275) · discourse/discourse@cfd5078 · GitHub

    • 帖子中的 URL 有时会被错误地分词并重复索引，导致链接内的词汇排名偏高。
    • 帖子中的灯箱内容会将其图像元数据污染到搜索结果中。

  3. 空帖子（如版主操作或小型帖子操作，例如关闭、分配）不再包含在搜索索引中。这一变化使索引更小，并减少了搜索结果中的噪音。FIX: Don't index posts with empty `Post#raw` for search. (#7263) · discourse/discourse@daeda80 · GitHub

  4. 通过从索引中删除已丢弃主题的帖子，缩小了搜索索引。PERF: Delete search data of posts from trashed topics periodically. (… · discourse/discourse@d151425 · GitHub

  5. 当帖子被移动到其他主题时，其搜索数据未得到更新，导致帖子错误地出现在匹配旧主题分类的搜索中。FIX: Reindex post for search when post is moved to a different topic. · discourse/discourse@d808f36 · GitHub

  6. 通过排除帖子的已渲染版本，减小了搜索结果的负载。PERF: Reduce number of queries and size of payload when searching. · discourse/discourse@03c6b22 · GitHub

  7. 在搜索精确短语时，搜索结果中帖子的摘要功能曾出现损坏，导致客户端缺少搜索词高亮。FIX: Post blurb incorrect when search contains a phrase match. · discourse/discourse@dae0bb4 · GitHub

请在这些更改之后告诉我们您的搜索结果是否有所改善。我们也希望了解您是否认为搜索结果变差了，以便我们继续优化。谢谢大家！

43 个赞
Discourse Version 2.3
Exclude category from search
AI Search for finding non-exact matches
Available settings within the category edit page
Search Priority for Categories don't work with empty search query?
4

Is the age of a post part of the weight in ranking search results? Information gets stale fairly quickly in our forum, so it would be nice to have a way to reduce the relevance of older postings without actually eliminating them.

11 个赞
5

Not yet, but it is an interesting idea, even in the weaker form of simply factoring in the date the topic was last touched

7 个赞
6

I would love to see a way to “pin” topics in search results so they appear first. That way common searches things can be lead directly to our tutorials.

7

You do this by changing search priority for the category those topics are in.

8

My “issue” is that those topics are spread through all forum categories on contextual basis. Hence I can’t use the category-based approach.

9

Then you are stuck because there is no random per topic way to do this, and there never will be.

10

Maybe search priorities for tags could be a thing eventually? Although not as straight forward since topics can have multiple tags…

Being able to weight categories is a great feature anyway! I already have several spammy special-use categories on my forum that will be nice to push down a bit in search.

6 个赞
Feature Request: Sponsored Posts
11

Yes, I see this happening at some point.

10 个赞
12

That would be awesome! That way blog posts can be tagged as such and prioritized for search…

1 个赞
13

I would love to see a way for certain tags to affect the search priority also.

4 个赞
14

有计划将其添加到 2.4.X 或 2.5 版本吗？

15

这是现在的趋势吗？

我没有足够的知识来解读这个资源：Search Controller Need help with understanding how discourse search works - #3 by neounix

2 个赞
17

还没有。目前只有类别具有可配置的搜索优先级。

3 个赞