From ebcf13c7d58463b0f3fc08330f29995ca0c48075 Mon Sep 17 00:00:00 2001 From: Andrew Kane Date: Sun, 14 Jul 2013 20:39:31 -0700 Subject: [PATCH] Updated readme --- README.md | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/README.md b/README.md index 948b40b..4e793db 100644 --- a/README.md +++ b/README.md @@ -6,7 +6,7 @@ Search made easy ### Searches -Searchkick provides sensible search defaults out of the box. It handles: +Searchkick provides sensible search defaults out of the box. It handles: - stemming - `tomatoes` matches `tomato` - special characters - `jalapenos` matches `jalapeños` @@ -15,7 +15,9 @@ Searchkick provides sensible search defaults out of the box. It handles: ### Make Searches Better Over Time -Use analytics on search conversions to improve results +Use analytics on search conversions to improve results. + +Also, give popular documents a little boost. ### Zero Downtime Changes @@ -57,9 +59,9 @@ When changing the mapping in a model, you must create a new index for the change curl http://localhost:9200/books/_mapping ``` -### Low Number of Documents +### Inconsistent Scores -By default, Tire creates an index on 5 shards - even in development. With a low number of documents, you will get inconsistent relevance scores by default. You can [read more about it here](http://www.elasticsearch.org/blog/understanding-query-then-fetch-vs-dfs-query-then-fetch/). To fix this, set the search type to `dfs_query_and_fetch`. Alternatively, you can just use one shard `settings: {number_of_shards: 1}`. +Elasticsearch was built to be distributed. However, with a low number of documents, the default settings can give us incorrect results. You can [read more about it here](http://www.elasticsearch.org/blog/understanding-query-then-fetch-vs-dfs-query-then-fetch/). To fix this, set the search type to `dfs_query_and_fetch`. Alternatively, you can just use one shard with `settings: {number_of_shards: 1}`. ## Installation -- libgit2 0.21.0