Copyright | (c) CNRS 2017-Present |
---|---|
License | AGPL + CECILL v3 |
Maintainer | team@gargantext.org |
Stability | experimental |
Portability | POSIX |
Safe Haskell | Safe-Inferred |
Language | Haskell2010 |
Synopsis
- goodMapListSize :: Int
- buildNgramsLists :: (HasNodeStory env err m, HasNLPServer env, HasTreeError err, HasNodeError err) => User -> UserCorpusId -> MasterCorpusId -> Maybe FlowSocialListWith -> GroupParams -> m (Map NgramsType [NgramsElement])
- newtype MapListSize = MapListSize {
- unMapListSize :: Int
- newtype MaxListSize = MaxListSize {
- unMaxListSize :: Int
- buildNgramsOthersList :: (HasNodeError err, HasNLPServer env, HasNodeStory env err m, HasTreeError err) => User -> UserCorpusId -> Maybe FlowSocialListWith -> GroupParams -> (NgramsType, MapListSize, MaxListSize) -> m (Map NgramsType [NgramsElement])
- getGroupParams :: (HasNodeError err, HasTreeError err) => GroupParams -> HashSet Ngrams -> DBCmd err GroupParams
- buildNgramsTermsList :: (HasNodeError err, HasNLPServer env, HasNodeStory env err m, HasTreeError err) => User -> UserCorpusId -> MasterCorpusId -> Maybe FlowSocialListWith -> GroupParams -> (NgramsType, MapListSize) -> m (Map NgramsType [NgramsElement])
Documentation
goodMapListSize :: Int #
Good value from users' requests and anthropological analysis
buildNgramsLists :: (HasNodeStory env err m, HasNLPServer env, HasTreeError err, HasNodeError err) => User -> UserCorpusId -> MasterCorpusId -> Maybe FlowSocialListWith -> GroupParams -> m (Map NgramsType [NgramsElement]) #
TODO improve grouping functions of Authors, Sources, Institutes..
newtype MapListSize #
newtype MaxListSize #
buildNgramsOthersList :: (HasNodeError err, HasNLPServer env, HasNodeStory env err m, HasTreeError err) => User -> UserCorpusId -> Maybe FlowSocialListWith -> GroupParams -> (NgramsType, MapListSize, MaxListSize) -> m (Map NgramsType [NgramsElement]) #
getGroupParams :: (HasNodeError err, HasTreeError err) => GroupParams -> HashSet Ngrams -> DBCmd err GroupParams #
https://gitlab.iscpif.fr/gargantext/haskell-gargantext/issues/169#note_10049 Stemming can be useful if you do not have any context: ok for full text search then.
In document, we have context so we can add grammar and linguistics rules to be more precise than the stemmatization, that is why the lemmatization is used here to group. Basically it will avoid grouping homonyms in list. In search usually you add more context to "control" the stemmatization approximation.
buildNgramsTermsList :: (HasNodeError err, HasNLPServer env, HasNodeStory env err m, HasTreeError err) => User -> UserCorpusId -> MasterCorpusId -> Maybe FlowSocialListWith -> GroupParams -> (NgramsType, MapListSize) -> m (Map NgramsType [NgramsElement]) #