[
https://issues.apache.org/jira/browse/LUCENE-5350?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13842194#comment-13842194
]
Areek Zillur commented on LUCENE-5350:
--------------------------------------
The Benchmark results (similar format to LookupBenchmark results) were as
follows:
{code}
-- input stats
Input size: 53022, numContexts: 2666, Avg. Input/Context: 20
-- construction time
AnalyzingSuggester input: 53022, time[ms]: 10140 [+- 149.13]
ContextAwareAnalyzingSuggester input: 53022, time[ms]: 1683 [+- 17.54]
-- RAM consumption
AnalyzingSuggester size[B]: 4,675,508
ContextAwareAnalyzingSuggester size[B]: 4,837,187
-- prefixes: 6-9, num: 7, onlyMorePopular: false
ContextAwareAnalyzingSuggester queries: 53022, time[ms]: 1277 [+- 14.67],
~kQPS: 42
AnalyzingSuggester queries: 53022, time[ms]: 2269 [+- 152.00], ~kQPS: 23
-- prefixes: 2-4, num: 7, onlyMorePopular: false
ContextAwareAnalyzingSuggester queries: 53022, time[ms]: 2294 [+- 24.02],
~kQPS: 23
AnalyzingSuggester queries: 53022, time[ms]: 4947 [+- 90.36], ~kQPS: 11
-- prefixes: 100-200, num: 7, onlyMorePopular: false
ContextAwareAnalyzingSuggester queries: 53022, time[ms]: 1177 [+- 32.13],
~kQPS: 45
AnalyzingSuggester queries: 53022, time[ms]: 935 [+- 11.53], ~kQPS: 57
{code}
>From the results it seems
ContextAwaresuggester compared to AnalyzingSuggester:
- is 6 times faster for construction
- consumes ~3% more RAM
- has ~ 2 times the QPS [for prefixes 6-9 & 2-4]
- has 20% less QPS [for prefix 100-200]
It is to be noted that the dataset only contains terms with one context (hence
the benchmark does not take into account terms with multiple contexts).
This was an interesting benchmark, thoughts?
> Add Context Aware Suggester
> ---------------------------
>
> Key: LUCENE-5350
> URL: https://issues.apache.org/jira/browse/LUCENE-5350
> Project: Lucene - Core
> Issue Type: New Feature
> Components: core/search
> Reporter: Areek Zillur
> Fix For: 5.0, 4.7
>
> Attachments: LUCENE-5350-benchmark.patch, LUCENE-5350.patch
>
>
> It would be nice to have a Context Aware Suggester (i.e. a suggester that
> could return suggestions depending on some specified context(s)).
> Use-cases:
> - location-based suggestions:
> -- returns suggestions which 'match' the context of a particular area
> --- suggest restaurants names which are in Palo Alto (context ->
> Palo Alto)
> - category-based suggestions:
> -- returns suggestions for items that are only in certain
> categories/genres (contexts)
> --- suggest movies that are of the genre sci-fi and adventure
> (context -> [sci-fi, adventure])
--
This message was sent by Atlassian JIRA
(v6.1#6144)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
