1,071
edits
Update:Remora:Search: Difference between revisions
Jump to navigation
Jump to search
→Relevant AMO Pages/Samples
| Line 12: | Line 12: | ||
* {{bug|439227}} includes some suggestions on what a good natural order might be | * {{bug|439227}} includes some suggestions on what a good natural order might be | ||
* {{bug|291413}} include number of ratings in the ranking | * {{bug|291413}} include number of ratings in the ranking | ||
* {{bug|378657}} exact names should sort first | |||
== Ordering Parameters == | == Ordering Parameters == | ||
There is a lot of data to work with to help order the output. These include: | There is potentially a lot of data to work with to help order the list output. These include: | ||
* Name | * Add-on Name/Title | ||
* Description | * Description | ||
* Rating | * Rating | ||
| Line 25: | Line 26: | ||
== Basic search == | == Basic search == | ||
Based on the most popular search terms seen above, we try to infer the intent of what is being searched for. (We may be wrong but we can at least guess the intent). It appears that searches cluster into the following categories: | |||
* Exact Name, e.g. "firebug", "ie tab" - clearly users are looking for these named add-ons, so the algo should bias to display these at the top | |||
* Functionality, e.g. weather | |||
* Web site, e.g. google, gmail | |||
see different types of basic searches being | |||
* Use the "natural order" algorithm | * Use the "natural order" algorithm | ||
* Do not use AMO ID sorting, it biases for longevity on the AMO site. Not necessarily on relevance. | * Do not use AMO ID sorting, it biases for longevity on the AMO site. Not necessarily on relevance. | ||