jlemoine

16 ms·

121.
by jlemoine 13y ago
By traditional approach we means all engines that use a unique score to rank documents (like all engine based on Lucene). SQL queries are not relevant for text query, for example you have no notion of tokenization or proximity between words
122.
by jlemoine 13y ago
You can try to search-as-you-type on our hacker news search to see the difference with other search engines: http://hn.algolia.com/ You have relevant results after each keystroke, even with typos. Classical engines use appr
123.
by jlemoine 13y ago
200ms does not allow you to provide search-as-you type and our e-commerce customers see that performance is linked to conversion. In term of relevance, we have developed a ranking algorithm specialized on that case that provide better resul
124.
by jlemoine 13y ago
Swiftype is a great tool to search for webpages. Our focus is to search in a database (for example products in e-commerce website, persons in a social network or CRM, ...). We offer features dedicated for that use case, you can have a look
125.
by jlemoine 13y ago
The search world is very big :) 5 mil records is nothing if you index logs (which is not Algolia typical use-cases) but for example this is big from an e-commerce perspective.
126.
by jlemoine 13y ago
(I am the Co-founder of Algolia) Thanks for your comment, you are right that the data are not perfectly indexed for this query. We have taken data from one of our customer and his use case is to search only for TV show names. We will improv
127.
Show HN: instant search on kickstarter(demos.algolia.com)
2 points by jlemoine 13y ago|0 comments
128.
Algolia (YC W14) How we made our onboarding frictionless(blog.algolia.com)
2 points by jlemoine 13y ago|0 comments
129.
Postmortem of today's 8 min indexing downtime(blog.algolia.com)
3 points by jlemoine 13y ago|0 comments
130.
by jlemoine 13y ago
Search queries on news.ycombinator.com are redirected to hn.algolia.io which is an alias for hn.algolia.com. Btw we have just put in production a new version with sort by date :)
131.
by jlemoine 13y ago
We took them into account and updated it. So comes this thread, to ask feedback about the second version.
132.
by jlemoine 13y ago
We already search in url (after title and content)
133.
Ask HN: What do you think about our last HN Search update?
28 points by jlemoine 13y ago|27 comments
134.
by jlemoine 13y ago
The next iteration is planned for tomorrow, we are taking into account the maximum of feature request possible. We will ask for new feedback after :)
135.
by jlemoine 13y ago
We will make changes in the next couple of days and we will ask feedback again. We prefer to launch sooner with few features and get community feedback rather than developing tons of features that we are the only to use :)
136.
by jlemoine 13y ago
This is the first iteration, we will improve it with your feedback
137.
by jlemoine 13y ago
Thanks for your feedback, we are collecting missing features and will improve it.
138.
by jlemoine 13y ago
Yes this is a real "update", we explain details on our blog post: http://blog.algolia.com/hacker-news-search-algolia/
139.
by jlemoine 13y ago
No yet, we have only indexed text. We will collect feedback to improve it, create_at will be part of the next update.
140.
by jlemoine 13y ago
The set of features is different with for example typo tolerance, if the limit of 1000 entries is a problem, we can increase it. We have written a blog post about the change if you want more details: http://blog.algolia.com/
141.
by jlemoine 13y ago
Yes for the moment this is manually reviewed, we will try to make it fully automatic but this is a very complex topic. We will write a blog post that explain the whole ranking.
142.
by jlemoine 13y ago
It is the Search Grader that is new, we have just released it today.
143.
Show HN: Search Grader. Check out Amazon, Craigslist & others' scores(grader.algolia.com)
17 points by jlemoine 13y ago|5 comments
144.
by jlemoine 13y ago
This approach is far from been perfect. A simple recursive implementation of Levenshtein on a Trie representing the dictionary would produce better performance!
145.
Discounts on 12 SaaS products for startups(rudebaguette.com)
1 points by jlemoine 13y ago|0 comments
146.
by jlemoine 13y ago
Firepad.io is missing from the list and has several derivate projects listed on their homepage.
147.
Show HN: How I added typeahead search in Jekyll blog with 50 lines of code(sylvain.utard.info)
4 points by jlemoine 13y ago|0 comments
148.
My “Must Have/Nice to Have” Framework(clementvouillon.com)
1 points by jlemoine 13y ago|0 comments
149.
Algolia Search API raises $1.5M seed round(pandodaily.com)
2 points by jlemoine 13y ago|0 comments
150.
by jlemoine 13y ago
Thanks for your feedback, I am the author of the article. 1) You are right that it is possible to mix both popularity and relevance, but you need to use boost and store everything in the float _score. This is dangerous and has side effects

More ›