An aggregation that returns interesting or unusual occurrences of free-text terms in a set.
It is like the <<search-aggregations-bucket-significantterms-aggregation,significant terms>> aggregation but differs in that:
* It is specifically designed for use on type `text` fields
* It does not require field data or doc-values
* It re-analyzes text content on-the-fly meaning it can also filter duplicate sections of
noisy text that otherwise tend to skew statistics.
WARNING: Re-analyzing _large_ result sets will require a lot of time and memory. It is recommended that the significant_text
aggregation is used as a child of either the <<search-aggregations-bucket-sampler-aggregation,sampler>> or
<<search-aggregations-bucket-diversified-sampler-aggregation,diversified sampler>> aggregation to limit the analysis
to a _small_ selection of top-matching documents e.g. 200. This will typically improve speed, memory use and quality of
results.
.Example use cases:
* Suggesting "H5N1" when users search for "bird flu" to help expand queries
* Suggesting keywords relating to stock symbol $ATI for use in an automated news classifier
In these cases the words being selected are not simply the most popular terms in results. The most popular words tend to be
very boring (_and, of, the, we, I, they_ ...).
The significant words are the ones that have undergone a significant change in popularity measured between a _foreground_ and _background_ set.
If the term "H5N1" only exists in 5 documents in a 10 million document index and yet is found in 4 of the 100 documents that make up a user's search results
that is significant and probably very relevant to their search. 5/10,000,000 vs 4/100 is a big swing in frequency.
experimental[The `significant_text` aggregation is new and may change in non-backwards compatible ways if we add further text-analysis features e.g. phrase detection]
==== Basic use
In the typical use case, the _foreground_ set of interest is a selection of the top-matching search results for a query
and the _background_set used for statistical comparisons is the index or indices from which the results were gathered.
Mr Pozmantier and other one-off associations with elasticsearch no longer appear in the aggregation
results as a consequence of copy-and-paste operations or other forms of mechanical repetition.
If your duplicate or near-duplicate content is identifiable via a single-value indexed field (perhaps
a hash of the article's `title` text or an `original_press_release_url` field) then it would be more
efficient to use a parent <<search-aggregations-bucket-diversified-sampler-aggregation,diversified sampler>> aggregation
to eliminate these documents from the sample set based on that single key. The less duplicate content you can feed into
the significant_text aggregation up front the better in terms of performance.
.How are the significance scores calculated?
**********************************
The numbers returned for scores are primarily intended for ranking different suggestions sensibly rather than something easily
understood by end users. The scores are derived from the doc frequencies in _foreground_ and _background_ sets. In brief, a
term is considered significant if there is a noticeable difference in the frequency in which a term appears in the subset and
in the background. The way the terms are ranked can be configured, see "Parameters" section.
**********************************
.Use the _"like this but not this"_ pattern
**********************************
You can spot mis-categorized content by first searching a structured field e.g. `category:adultMovie` and use significant_text on the
text "movie_description" field. Take the suggested words (I'll leave them to your imagination) and then search for all movies NOT marked as category:adultMovie but containing these keywords.
You now have a ranked list of badly-categorized movies that you should reclassify or at least remove from the "familyFriendly" category.
The significance score from each term can also provide a useful `boost` setting to sort matches.
Using the `minimum_should_match` setting of the `terms` query with the keywords will help control the balance of precision/recall in the result set i.e
a high setting would have a small number of relevant results packed full of keywords and a setting of "1" would produce a more exhaustive results set with all documents containing _any_ keyword.
**********************************
==== Limitations
===== No support for child aggregations
The significant_text aggregation intentionally does not support the addition of child aggregations because:
* It would come with a high memory cost
* It isn't a generally useful feature and there is a workaround for those that need it
The volume of candidate terms is generally very high and these are pruned heavily before the final
results are returned. Supporting child aggregations would generate additional churn and be inefficient.
Clients can always take the heavily-trimmed set of results from a `significant_text` request and
make a subsequent follow-up query using a `terms` aggregation with an `include` clause and child
aggregations to perform further analysis of selected keywords in a more efficient fashion.
The counts of how many documents contain a term provided in results are based on summing the samples returned from each shard and
as such may be:
* low if certain shards did not provide figures for a given term in their top sample
* high when considering the background frequency as it may count occurrences found in deleted documents
Like most design decisions, this is the basis of a trade-off in which we have chosen to provide fast performance at the cost of some (typically small) inaccuracies.
However, the `size` and `shard size` settings covered in the next section provide tools to help control the accuracy levels.
==== Parameters
===== Significance heuristics
This aggregation supports the same scoring heuristics (JLH, mutual_information, gnd, chi_square etc) as the <<search-aggregations-bucket-significantterms-aggregation,significant terms>> aggregation
===== Size & Shard Size
The `size` parameter can be set to define how many term buckets should be returned out of the overall terms list. By
default, the node coordinating the search process will request each shard to provide its own top term buckets
and once all shards respond, it will reduce the results to the final list that will then be returned to the client.
If the number of unique terms is greater than `size`, the returned list can be slightly off and not accurate
(it could be that the term counts are slightly off and it could even be that a term that should have been in the top
size buckets was not returned).
To ensure better accuracy a multiple of the final `size` is used as the number of terms to request from each shard
using a heuristic based on the number of shards. To take manual control of this setting the `shard_size` parameter
can be used to control the volumes of candidate terms produced by each shard.
Low-frequency terms can turn out to be the most interesting ones once all results are combined so the
significant_terms aggregation can produce higher-quality results when the `shard_size` parameter is set to
values significantly higher than the `size` setting. This ensures that a bigger volume of promising candidate terms are given
a consolidated review by the reducing node before the final selection. Obviously large candidate term lists
will cause extra network traffic and RAM usage so this is quality/cost trade off that needs to be balanced. If `shard_size` is set to -1 (the default) then `shard_size` will be automatically estimated based on the number of shards and the `size` parameter.
NOTE: `shard_size` cannot be smaller than `size` (as it doesn't make much sense). When it is, elasticsearch will
override it and reset it to be equal to `size`.
===== Minimum document count
It is possible to only return terms that match more than a configured number of hits using the `min_doc_count` option.
The Default value is 3.
Terms that score highly will be collected on a shard level and merged with the terms collected from other shards in a second step.
However, the shard does not have the information about the global term frequencies available. The decision if a term is added to a
candidate list depends only on the score computed on the shard using local shard frequencies, not the global frequencies of the word.
The `min_doc_count` criterion is only applied after merging local terms statistics of all shards. In a way the decision to add the
term as a candidate is made without being very _certain_ about if the term will actually reach the required `min_doc_count`.
This might cause many (globally) high frequent terms to be missing in the final result if low frequent but high scoring terms populated
the candidate lists. To avoid this, the `shard_size` parameter can be increased to allow more candidate terms on the shards.
However, this increases memory consumption and network traffic.
`shard_min_doc_count` parameter
The parameter `shard_min_doc_count` regulates the _certainty_ a shard has if the term should actually be added to the candidate list or
not with respect to the `min_doc_count`. Terms will only be considered if their local shard frequency within the set is higher than the
`shard_min_doc_count`. If your dictionary contains many low frequent words and you are not interested in these (for example misspellings),
then you can set the `shard_min_doc_count` parameter to filter out candidate terms on a shard level that will with a reasonable certainty
not reach the required `min_doc_count` even after merging the local frequencies. `shard_min_doc_count` is set to `1` per default and has
no effect unless you explicitly set it.
WARNING: Setting `min_doc_count` to `1` is generally not advised as it tends to return terms that
are typos or other bizarre curiosities. Finding more than one instance of a term helps
reinforce that, while still rare, the term was not the result of a one-off accident. The
default value of 3 is used to provide a minimum weight-of-evidence.
Setting `shard_min_doc_count` too high will cause significant candidate terms to be filtered out on a shard level.
This value should be set much lower than `min_doc_count/#shards`.
===== Custom background context
The default source of statistical information for background term frequencies is the entire index and this
scope can be narrowed through the use of a `background_filter` to focus in on significant terms within a narrower
