Finding You the Best Hotel at LateRooms.com with Elasticsearch

https://www.elastic.co/blog/finding-you-the-best-hotel-at-laterooms-com-with-elasticsearch

Good overview. Some of the ideas and approaches are the same as ours.

proper sorting is not that easy

Trying to sort site-wide search results on most e-commerce sites typically end up a mess, with irrelevant search results being propelled to the top of the list as users sort by price, customer ratings, etc

http://baymard.com/blog/faceted-sorting

Basic NLP with the bash https://gibrown.com/2013/01/26/unix-bi-grams-tri-grams-and-topic-modeling/

Bags of Words vs. Bags of Things

 What most users don’t know is that the search engine is answering a different question than they are asking – it returns the documents that have the words that they entered (or synonyms if the application designers have provided this) which is sometimes not exactly “what” they are looking for.

https://lucidworks.com/blog/2014/07/02/automatic-phrase-tokenization-improving-lucene-search-precision-by-more-precise-linguistic-analysis/

The user is telling the search engine what they want. They are looking for specific things, not specific words. The more that we can do to redress this mismatch between tokens and things, the better the user experience will be.

I've found new awesome blog on search - http://opensourceconnections.com/blog/

Some of the articles in that blog:

• Search Precision and Recall By Example
• Semantic Search with Latent Semantic Analysis
• Search Relevancy, is there a magic bullet?
• Bad Patterns in Tuning Search Results

And many other

Product Insights for Airbnb

Got a lot of fresh ideas how to improve our ranking model

lucene solr commit history 1