Time/EN: Fix "in" + year

Summary: Now that years are latent, let's absorb the "in" and make them not latent.

Reviewed By: zliu41

Differential Revision: D7587599

fbshipit-source-id: 61a19ac389244df491591d78c28f0301f9124439
This commit is contained in:
Julien Odent 2018-04-11 10:48:52 -07:00 committed by Facebook Github Bot
parent 528fea9ad4
commit 120f569ed0
25 changed files with 11269 additions and 7140 deletions

View File

@ -50,8 +50,13 @@ fullParses l sentence targets = debugTokens sentence .
where
n = Text.length sentence
debugCustom :: Context -> Options -> Text -> [Some Dimension] -> IO [Entity]
debugCustom context options sentence targets = debugTokens sentence .
analyze sentence context options $ HashSet.fromList targets
ptree :: Text -> Entity -> IO ()
ptree sentence Entity {enode} = pnode sentence 0 enode
-- -----------------------------------------------------------------
-- Internals
@ -65,10 +70,6 @@ parses l sentence targets = flip filter tokens $
tokens = parseAndResolve rules sentence testContext {locale = l} testOptions
rules = rulesFor l $ HashSet.fromList targets
debugCustom :: Context -> Options -> Text -> [Some Dimension] -> IO [Entity]
debugCustom context options sentence targets = debugTokens sentence .
analyze sentence context options $ HashSet.fromList targets
debugTokens :: Text -> [ResolvedToken] -> IO [Entity]
debugTokens sentence tokens = do
mapM_ (ptree sentence) entities

View File

@ -35,11 +35,11 @@ classifiers
likelihoods = HashMap.fromList [], n = 0}}),
("integer (numeric)",
Classifier{okData =
ClassData{prior = -0.9246589815446716, unseen = -4.795790545596741,
ClassData{prior = -0.921320080279157, unseen = -4.795790545596741,
likelihoods = HashMap.fromList [("", 0.0)], n = 119},
koData =
ClassData{prior = -0.5052854433903752, unseen = -5.209486152841421,
likelihoods = HashMap.fromList [("", 0.0)], n = 181}}),
ClassData{prior = -0.507486722500476, unseen = -5.204006687076795,
likelihoods = HashMap.fromList [("", 0.0)], n = 180}}),
("exactly <time-of-day>",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.3025850929940455,
@ -52,13 +52,6 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -1.3862943611198906,
likelihoods = HashMap.fromList [], n = 0}}),
("Father's Day",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("<hour-of-day> <half> (as relative minutes)",
Classifier{okData =
ClassData{prior = -infinity, unseen = -1.0986122886681098,
@ -216,6 +209,13 @@ classifiers
("<time-of-day> am|pm", -1.9095425048844386),
("hour", -1.349926716949016), ("minute", -2.1972245773362196)],
n = 8}}),
("Neujahr",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.3862943611198906,
likelihoods = HashMap.fromList [("", 0.0)], n = 2},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("absorption of , after named day",
Classifier{okData =
ClassData{prior = -7.410797215372185e-2,
@ -301,6 +301,13 @@ classifiers
[("hh:mmtime-of-day (latent)", -1.0986122886681098),
("minutehour", -1.0986122886681098)],
n = 1}}),
("Halloween",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("between <datetime> and <datetime> (interval)",
Classifier{okData =
ClassData{prior = -0.6931471805599453,
@ -365,6 +372,13 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("Valentinstag",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("hour (grain)",
Classifier{okData =
ClassData{prior = -0.15415067982725836,
@ -391,285 +405,286 @@ classifiers
n = 1}}),
("intersect",
Classifier{okData =
ClassData{prior = -0.2639842810219578, unseen = -6.92951677076365,
ClassData{prior = -0.26452876747706405,
unseen = -6.926577033222725,
likelihoods =
HashMap.fromList
[("Samstag<time-of-day> o'clock", -6.23539063760472),
("hourday", -4.982627669109352),
("<datetime> - <datetime> (interval)year", -4.982627669109352),
("dayhour", -2.571828991475074),
("daymonth", -3.167337702471103),
[("Samstag<time-of-day> o'clock", -6.2324480165505225),
("hourday", -4.979685048055154),
("<datetime> - <datetime> (interval)year", -4.979685048055154),
("dayhour", -2.568886370420876),
("daymonth", -3.1879255788270995),
("<day-of-month>(ordinal) <named-month> year<time-of-day> o'clock",
-6.23539063760472),
("monthyear", -4.5306425453662955),
("yearhour", -6.23539063760472),
("Juliintersect", -6.23539063760472),
("intersect<time-of-day> o'clock", -4.095324474108449),
("christmasyear", -6.23539063760472),
("after lunchat <time-of-day>", -6.23539063760472),
-6.2324480165505225),
("monthyear", -4.527699924312097),
("yearhour", -6.2324480165505225),
("Juliintersect", -6.2324480165505225),
("intersect<time-of-day> o'clock", -4.092381853054252),
("after lunchat <time-of-day>", -6.2324480165505225),
("absorption of , after named daymm/dd/yyyy",
-6.23539063760472),
("intersectJuli", -5.5422434570447745),
("<day-of-month> (ordinal)Dienstag", -6.23539063760472),
("intersect by 'of', 'from', 'syear", -5.829925529496556),
("<time-of-day> am|pmintersect", -5.136778348936611),
("<day-of-month> (ordinal)September", -5.829925529496556),
("intersect<time> <part-of-day>", -4.155949095924885),
("<time-of-day> o'clockafter lunch", -5.5422434570447745),
("<day-of-month> (ordinal)Mittwoch", -6.23539063760472),
("Montagthis <cycle>", -6.23539063760472),
-6.2324480165505225),
("intersectJuli", -5.539300835990577),
("<day-of-month> (ordinal)Dienstag", -6.2324480165505225),
("intersect by 'of', 'from', 'syear", -5.826982908442359),
("<time-of-day> am|pmintersect", -5.133835727882413),
("<day-of-month> (ordinal)September", -5.826982908442359),
("intersect<time> <part-of-day>", -4.153006474870687),
("<time-of-day> o'clockafter lunch", -5.539300835990577),
("<day-of-month> (ordinal)Mittwoch", -6.2324480165505225),
("Montagthis <cycle>", -6.2324480165505225),
("<time> <part-of-day><time-of-day> o'clock",
-6.23539063760472),
("Oktoberyear", -5.136778348936611),
("today<time-of-day> o'clock", -6.23539063760472),
("<time-of-day> o'clockon <date>", -5.829925529496556),
("<time-of-day> am|pmintersect by ','", -5.5422434570447745),
("intersect by ','year", -5.136778348936611),
("on <date><time-of-day> o'clock", -6.23539063760472),
("exactly <time-of-day>tomorrow", -5.829925529496556),
("mm/dd<time-of-day> o'clock", -6.23539063760472),
("monthhour", -5.829925529496556),
-6.2324480165505225),
("Oktoberyear", -5.133835727882413),
("today<time-of-day> o'clock", -6.2324480165505225),
("<time-of-day> o'clockon <date>", -5.826982908442359),
("<time-of-day> am|pmintersect by ','", -5.539300835990577),
("intersect by ','year", -5.133835727882413),
("on <date><time-of-day> o'clock", -6.2324480165505225),
("exactly <time-of-day>tomorrow", -5.826982908442359),
("mm/dd<time-of-day> o'clock", -6.2324480165505225),
("monthhour", -5.826982908442359),
("on <date>between <datetime> and <datetime> (interval)",
-6.23539063760472),
("last <day-of-week> of <time>year", -6.23539063760472),
("hourmonth", -5.829925529496556),
("todayat <time-of-day>", -5.5422434570447745),
("mm/ddabout <time-of-day>", -5.829925529496556),
-6.2324480165505225),
("last <day-of-week> of <time>year", -6.2324480165505225),
("hourmonth", -5.826982908442359),
("todayat <time-of-day>", -5.539300835990577),
("mm/ddabout <time-of-day>", -5.826982908442359),
("Donnerstagfrom <datetime> - <datetime> (interval)",
-5.829925529496556),
-5.826982908442359),
("on <date>between <time-of-day> and <time-of-day> (interval)",
-6.23539063760472),
("on <date>at <time-of-day>", -5.829925529496556),
("dayday", -3.0573368072567746),
("<time> <part-of-day>at <time-of-day>", -5.829925529496556),
-6.2324480165505225),
("on <date>at <time-of-day>", -5.826982908442359),
("dayday", -3.054394186202577),
("<time> <part-of-day>at <time-of-day>", -5.826982908442359),
("<time-of-day> am|pmabsorption of , after named day",
-6.23539063760472),
("about <time-of-day>on <date>", -6.23539063760472),
-6.2324480165505225),
("about <time-of-day>on <date>", -6.2324480165505225),
("<hour-of-day> <integer> (as relative minutes)in|during the <part-of-day>",
-5.829925529496556),
("<day-of-month> (ordinal)intersect", -5.136778348936611),
("hourhour", -3.7096619932964647),
("hh:mmintersect by ','", -4.982627669109352),
-5.826982908442359),
("<day-of-month> (ordinal)intersect", -5.133835727882413),
("hourhour", -3.7067193722422673),
("hh:mmintersect by ','", -4.979685048055154),
("Donnerstagfrom <time-of-day> - <time-of-day> (interval)",
-5.5422434570447745),
("Dienstagthis <cycle>", -6.23539063760472),
("<part-of-day> of <time>Februar", -5.829925529496556),
("dayyear", -3.317619905520441),
("Mittwochthis <cycle>", -6.23539063760472),
-5.539300835990577),
("Dienstagthis <cycle>", -6.2324480165505225),
("<part-of-day> of <time>Februar", -5.826982908442359),
("dayyear", -3.3146772844662435),
("Mittwochthis <cycle>", -6.2324480165505225),
("<time-of-day> o'clockin|during the <part-of-day>",
-5.136778348936611),
("tomorrow<time-of-day> o'clock", -6.23539063760472),
("<time-of-day> o'clocktomorrow", -5.5422434570447745),
-5.133835727882413),
("tomorrow<time-of-day> o'clock", -6.2324480165505225),
("<time-of-day> o'clocktomorrow", -5.539300835990577),
("<day-of-month>(ordinal) <named-month>year",
-5.829925529496556),
-5.826982908442359),
("half <integer> (german style hour-of-day)after lunch",
-6.23539063760472),
-6.2324480165505225),
("absorption of , after named day<day-of-month>(ordinal) <named-month>",
-4.982627669109352),
("hourminute", -5.5422434570447745),
("the <day-of-month> (ordinal)Juli", -6.23539063760472),
-4.979685048055154),
("hourminute", -5.539300835990577),
("the <day-of-month> (ordinal)Juli", -6.2324480165505225),
("on <date><day-of-month>(ordinal) <named-month>",
-5.829925529496556),
("minutemonth", -3.5612419881781916),
("minutehour", -3.984098838998225),
-5.826982908442359),
("minutemonth", -3.558299367123994),
("minutehour", -3.9811562179440276),
("at <time-of-day>in|during the <part-of-day>",
-5.319099905730565),
-5.316157284676367),
("absorption of , after named day<day-of-month>(ordinal) <named-month> year",
-5.829925529496556),
-5.826982908442359),
("absorption of , after named day<named-month> <day-of-month> (non ordinal)",
-5.136778348936611),
-5.133835727882413),
("<day-of-month>(ordinal) <named-month> year<time> <part-of-day>",
-6.23539063760472),
-6.2324480165505225),
("<hour-of-day> <integer> (as relative minutes)after lunch",
-6.23539063760472),
("Donnerstag<time> timezone", -5.829925529496556),
("Samstagat <time-of-day>", -5.829925529496556),
("<time-of-day> am|pmSamstag", -6.23539063760472),
-6.2324480165505225),
("Donnerstag<time> timezone", -5.826982908442359),
("Samstagat <time-of-day>", -5.826982908442359),
("<time-of-day> am|pmSamstag", -6.2324480165505225),
("intersect<day-of-month>(ordinal) <named-month>",
-4.84909627648483),
("this <part-of-day><time-of-day> o'clock", -6.23539063760472),
-4.846153655430633),
("this <part-of-day><time-of-day> o'clock",
-6.2324480165505225),
("<day-of-month>(ordinal) <named-month>intersect",
-6.23539063760472),
("hh:mmintersect", -3.793043602235516),
("Donnerstaghh:mm", -6.23539063760472),
("<day-of-month> (ordinal)Juli", -5.5422434570447745),
("intersect by ','intersect", -4.982627669109352),
("intersect by 'of', 'from', 'sJuli", -6.23539063760472),
("<day-of-month> (ordinal)Oktober", -6.23539063760472),
-6.2324480165505225),
("hh:mmintersect", -3.790100981181318),
("Donnerstaghh:mm", -6.2324480165505225),
("Weihnachtenyear", -6.2324480165505225),
("<day-of-month> (ordinal)Juli", -5.539300835990577),
("intersect by ','intersect", -4.979685048055154),
("intersect by 'of', 'from', 'sJuli", -6.2324480165505225),
("half <integer> (german style hour-of-day)in|during the <part-of-day>",
-5.829925529496556),
-5.826982908442359),
("from <datetime> - <datetime> (interval)year",
-5.5422434570447745),
("at <time-of-day>intersect", -4.443631168376665),
-5.539300835990577),
("at <time-of-day>intersect", -4.440688547322468),
("on <date>from <time-of-day> - <time-of-day> (interval)",
-5.829925529496556),
-5.826982908442359),
("<time> <part-of-day>from <time-of-day> - <time-of-day> (interval)",
-6.23539063760472),
-6.2324480165505225),
("absorption of , after named day<day-of-month> (ordinal)",
-4.982627669109352),
-4.979685048055154),
("Donnerstagbetween <time-of-day> and <time-of-day> (interval)",
-6.23539063760472),
("dayminute", -3.984098838998225),
-6.2324480165505225),
("dayminute", -3.9811562179440276),
("on <date>from <datetime> - <datetime> (interval)",
-6.23539063760472),
("Mittwochnext <cycle>", -6.23539063760472),
-6.2324480165505225),
("Mittwochnext <cycle>", -6.2324480165505225),
("Donnerstagbetween <datetime> and <datetime> (interval)",
-6.23539063760472),
-6.2324480165505225),
("<time> <part-of-day>from <datetime> - <datetime> (interval)",
-6.23539063760472),
("intersectyear", -4.443631168376665),
("on <date>intersect", -5.829925529496556),
("on <date><day-of-month> (ordinal)", -5.829925529496556),
("intersectSeptember", -3.4945506136795195),
("<ordinal> <cycle> of <time>year", -6.23539063760472),
("minuteday", -2.2557089837027595),
-6.2324480165505225),
("intersectyear", -4.440688547322468),
("on <date>intersect", -5.826982908442359),
("on <date><day-of-month> (ordinal)", -5.826982908442359),
("intersectSeptember", -3.4916079926253216),
("<ordinal> <cycle> of <time>year", -6.2324480165505225),
("minuteday", -2.252766362648562),
("absorption of , after named dayintersect",
-3.8374953648063497),
("intersect by ','September", -4.731313240828446),
("year<time-of-day> o'clock", -6.23539063760472),
("Juliyear", -5.829925529496556),
("at <time-of-day>intersect by ','", -5.5422434570447745),
("hh:mmabsorption of , after named day", -5.829925529496556),
("intersect by ','<time> <part-of-day>", -5.319099905730565),
("hh:mmon <date>", -3.793043602235516),
-3.8345527437521523),
("intersect by ','September", -4.728370619774249),
("year<time-of-day> o'clock", -6.2324480165505225),
("Juliyear", -5.826982908442359),
("at <time-of-day>intersect by ','", -5.539300835990577),
("hh:mmabsorption of , after named day", -5.826982908442359),
("intersect by ','<time> <part-of-day>", -5.316157284676367),
("hh:mmon <date>", -3.790100981181318),
("at <time-of-day>absorption of , after named day",
-6.23539063760472),
("until <time-of-day>after lunch", -6.23539063760472),
("mm/ddyear", -5.829925529496556),
("intersect by ','<time-of-day> o'clock", -5.319099905730565),
("intersect<day-of-month> (ordinal)", -4.84909627648483),
("absorption of , after named daymm/dd", -6.23539063760472),
("on <date>September", -5.319099905730565),
("Septemberyear", -5.829925529496556),
-6.2324480165505225),
("until <time-of-day>after lunch", -6.2324480165505225),
("mm/ddyear", -5.826982908442359),
("intersect by ','<time-of-day> o'clock", -5.316157284676367),
("intersect<day-of-month> (ordinal)", -4.846153655430633),
("absorption of , after named daymm/dd", -6.2324480165505225),
("on <date>September", -5.316157284676367),
("Septemberyear", -5.826982908442359),
("<day-of-month> (ordinal)intersect by 'of', 'from', 's",
-5.5422434570447745),
("at <time-of-day>on <date>", -4.363588460703129),
("intersectintersect", -4.095324474108449),
("<day-of-month> (ordinal)Februar", -5.136778348936611),
("dayweek", -5.136778348936611),
("intersect by ','Juli", -5.5422434570447745),
-5.539300835990577),
("at <time-of-day>on <date>", -4.360645839648932),
("intersectintersect", -4.092381853054252),
("<day-of-month> (ordinal)Februar", -5.133835727882413),
("dayweek", -5.133835727882413),
("intersect by ','Juli", -5.539300835990577),
("absorption of , after named daythe <day-of-month> (ordinal)",
-6.23539063760472),
("weekyear", -5.829925529496556),
("hh:mmin|during the <part-of-day>", -5.319099905730565),
("Marzyear", -6.23539063760472),
("tomorrowat <time-of-day>", -5.829925529496556),
-6.2324480165505225),
("weekyear", -5.826982908442359),
("hh:mmin|during the <part-of-day>", -5.316157284676367),
("Marzyear", -6.2324480165505225),
("tomorrowat <time-of-day>", -5.826982908442359),
("<hour-of-day> <integer> (as relative minutes)on <date>",
-6.23539063760472),
-6.2324480165505225),
("Donnerstag<time-of-day> - <time-of-day> (interval)",
-6.23539063760472),
-6.2324480165505225),
("Donnerstag<datetime> - <datetime> (interval)",
-6.23539063760472),
("Sonntaglast <cycle>", -6.23539063760472),
("at <time-of-day>tomorrow", -6.23539063760472),
-6.2324480165505225),
("Sonntaglast <cycle>", -6.2324480165505225),
("at <time-of-day>tomorrow", -6.2324480165505225),
("about <time-of-day>in|during the <part-of-day>",
-5.829925529496556),
-5.826982908442359),
("half <integer> (german style hour-of-day)on <date>",
-6.23539063760472),
("this <part-of-day>at <time-of-day>", -5.829925529496556),
("<datetime> - <datetime> (interval)Juli", -5.829925529496556),
-6.2324480165505225),
("this <part-of-day>at <time-of-day>", -5.826982908442359),
("<datetime> - <datetime> (interval)Juli", -5.826982908442359),
("after lunch<hour-of-day> <integer> (as relative minutes)",
-5.829925529496556),
("<day-of-month> (ordinal)Marz", -6.23539063760472),
("last <cycle> of <time>year", -5.829925529496556),
-5.826982908442359),
("<day-of-month> (ordinal)Marz", -6.2324480165505225),
("last <cycle> of <time>year", -5.826982908442359),
("<named-month> <day-of-month> (non ordinal)year",
-6.23539063760472),
-6.2324480165505225),
("<day-of-month> (non ordinal) <named-month>year",
-6.23539063760472)],
n = 427},
-6.2324480165505225)],
n = 426},
koData =
ClassData{prior = -1.4609558898889103, unseen = -6.05443934626937,
ClassData{prior = -1.4591557093847622, unseen = -6.052089168924417,
likelihoods =
HashMap.fromList
[("Juli<day-of-month> (non ordinal) <named-month>",
-5.358941988364472),
-5.356586274672012),
("absorption of , after named daythe <day-of-month> (non ordinal)",
-5.358941988364472),
("<datetime> - <datetime> (interval)year", -5.358941988364472),
("dayhour", -4.665794807804526),
-5.356586274672012),
("<datetime> - <datetime> (interval)year", -5.356586274672012),
("dayhour", -4.663439094112067),
("<named-month> <day-of-month> (non ordinal)Juli",
-5.358941988364472),
("daymonth", -3.48713981146288),
("monthday", -4.953476880256307),
("monthyear", -5.358941988364472),
("yearhour", -5.358941988364472),
("after lunchat <time-of-day>", -4.953476880256307),
("mm/dduntil <time-of-day>", -4.953476880256307),
("until <time-of-day>year", -4.953476880256307),
("<day-of-month> (ordinal)Dienstag", -4.953476880256307),
("absorption of , after named dayFebruar", -4.260329699696362),
("on <date>Februar", -5.358941988364472),
("intersect by 'of', 'from', 'syear", -4.665794807804526),
("<time-of-day> am|pmintersect", -3.654193896126046),
("intersect<time> <part-of-day>", -5.358941988364472),
("<time-of-day> o'clockafter lunch", -5.358941988364472),
("<day-of-month> (ordinal)Mittwoch", -5.358941988364472),
("after lunch<time-of-day> o'clock", -5.358941988364472),
("absorption of , after named dayhh:mm", -5.358941988364472),
("<time-of-day> am|pmintersect by ','", -4.106179019869104),
-5.356586274672012),
("daymonth", -3.484784097770421),
("monthday", -4.951121166563848),
("monthyear", -5.356586274672012),
("yearhour", -5.356586274672012),
("after lunchat <time-of-day>", -4.951121166563848),
("mm/dduntil <time-of-day>", -4.951121166563848),
("until <time-of-day>year", -4.951121166563848),
("<day-of-month> (ordinal)Dienstag", -4.951121166563848),
("absorption of , after named dayFebruar", -4.257973986003902),
("on <date>Februar", -5.356586274672012),
("intersect by 'of', 'from', 'syear", -4.663439094112067),
("<time-of-day> am|pmintersect", -3.651838182433587),
("intersect<time> <part-of-day>", -5.356586274672012),
("<time-of-day> o'clockafter lunch", -5.356586274672012),
("<day-of-month> (ordinal)Mittwoch", -5.356586274672012),
("after lunch<time-of-day> o'clock", -5.356586274672012),
("absorption of , after named dayhh:mm", -5.356586274672012),
("<time-of-day> am|pmintersect by ','", -4.103823306176644),
("<time-of-day> o'clock<time> <part-of-day>",
-4.442651256490317),
("<time-of-day> am|pmFreitag", -4.953476880256307),
("mm/ddhh:mm", -4.442651256490317),
("monthhour", -5.358941988364472),
-4.440295542797857),
("<time-of-day> am|pmFreitag", -4.951121166563848),
("mm/ddhh:mm", -4.440295542797857),
("monthhour", -5.356586274672012),
("on <date>between <datetime> and <datetime> (interval)",
-5.358941988364472),
("todayat <time-of-day>", -5.358941988364472),
-5.356586274672012),
("todayat <time-of-day>", -5.356586274672012),
("Donnerstagfrom <datetime> - <datetime> (interval)",
-4.953476880256307),
-4.951121166563848),
("on <date>between <time-of-day> and <time-of-day> (interval)",
-5.358941988364472),
("dayday", -3.972647627244581),
-5.356586274672012),
("dayday", -3.970291913552122),
("<time-of-day> am|pmabsorption of , after named day",
-4.953476880256307),
("hourhour", -3.8548645915881976),
-4.951121166563848),
("hourhour", -3.8525088778957386),
("Donnerstagfrom <time-of-day> - <time-of-day> (interval)",
-4.665794807804526),
("Donnerstagafter <time-of-day>", -5.358941988364472),
("dayyear", -3.2795004466846356),
("minutemonth", -3.48713981146288),
("Aprilyear", -5.358941988364472),
("Dienstagafter <time-of-day>", -5.358941988364472),
-4.663439094112067),
("Donnerstagafter <time-of-day>", -5.356586274672012),
("dayyear", -3.2771447329921766),
("minutemonth", -3.484784097770421),
("Aprilyear", -5.356586274672012),
("Dienstagafter <time-of-day>", -5.356586274672012),
("intersect<day-of-month>(ordinal) <named-month>",
-4.665794807804526),
("Donnerstaghh:mm", -4.953476880256307),
("<day-of-month> (ordinal)Juli", -4.953476880256307),
("<day-of-month> (ordinal)hh:mm", -4.442651256490317),
-4.663439094112067),
("Donnerstaghh:mm", -4.951121166563848),
("<day-of-month> (ordinal)Juli", -4.951121166563848),
("<day-of-month> (ordinal)hh:mm", -4.440295542797857),
("<named-month> <day-of-month> (non ordinal)August",
-5.358941988364472),
-5.356586274672012),
("August<day-of-month> (non ordinal) <named-month>",
-5.358941988364472),
("at <time-of-day>intersect", -4.260329699696362),
-5.356586274672012),
("at <time-of-day>intersect", -4.257973986003902),
("on <date>from <time-of-day> - <time-of-day> (interval)",
-4.953476880256307),
-4.951121166563848),
("Donnerstagbetween <time-of-day> and <time-of-day> (interval)",
-5.358941988364472),
("dayminute", -2.6847933389379426),
-5.356586274672012),
("dayminute", -2.6824376252454836),
("on <date>from <datetime> - <datetime> (interval)",
-5.358941988364472),
-5.356586274672012),
("Donnerstagbetween <datetime> and <datetime> (interval)",
-5.358941988364472),
("until <time-of-day>Juli", -5.358941988364472),
("intersectSeptember", -3.7495040759303713),
("minuteday", -2.3632097148104805),
("intersect by ','September", -4.665794807804526),
("at <time-of-day>intersect by ','", -4.665794807804526),
("at <time-of-day>Freitag", -5.358941988364472),
-5.356586274672012),
("until <time-of-day>Juli", -5.356586274672012),
("intersectSeptember", -3.747148362237912),
("minuteday", -2.3608540011180215),
("intersect by ','September", -4.663439094112067),
("at <time-of-day>intersect by ','", -4.663439094112067),
("at <time-of-day>Freitag", -5.356586274672012),
("at <time-of-day>absorption of , after named day",
-5.358941988364472),
("year<time> <part-of-day>", -5.358941988364472),
("mm/ddyear", -3.8548645915881976),
("intersect<day-of-month> (ordinal)", -4.665794807804526),
-5.356586274672012),
("year<time> <part-of-day>", -5.356586274672012),
("mm/ddyear", -3.8525088778957386),
("intersect<day-of-month> (ordinal)", -4.663439094112067),
("<day-of-month> (ordinal)intersect by 'of', 'from', 's",
-4.953476880256307),
("intersectintersect", -4.665794807804526),
("Donnerstag<part-of-day> of <time>", -5.358941988364472),
("<day-of-month> (ordinal)April", -5.358941988364472),
-4.951121166563848),
("intersectintersect", -4.663439094112067),
("Donnerstag<part-of-day> of <time>", -5.356586274672012),
("<day-of-month> (ordinal)April", -5.356586274672012),
("Donnerstag<time-of-day> - <time-of-day> (interval)",
-5.358941988364472),
("after <time-of-day>year", -5.358941988364472),
("on <date>after <time-of-day>", -5.358941988364472),
("tomorrownoon", -5.358941988364472)],
-5.356586274672012),
("after <time-of-day>year", -5.356586274672012),
("on <date>after <time-of-day>", -5.356586274672012),
("tomorrownoon", -5.356586274672012)],
n = 129}}),
("<ordinal> <cycle> of <time>",
Classifier{okData =
@ -730,6 +745,13 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("Muttertag",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("next <cycle>",
Classifier{okData =
ClassData{prior = -0.15415067982725836,
@ -778,6 +800,13 @@ classifiers
[("hh:mmtime-of-day (latent)", -1.0986122886681098),
("minutehour", -1.0986122886681098)],
n = 3}}),
("Silvester",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("year (latent)",
Classifier{okData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
@ -828,6 +857,13 @@ classifiers
ClassData{prior = -2.4849066497880004,
unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1}}),
("Weihnachten",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.791759469228055,
likelihoods = HashMap.fromList [("", 0.0)], n = 4},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("quarter to|till|before <integer> (hour-of-day)",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.6094379124341003,
@ -881,6 +917,13 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("Vatertag",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("week (grain)",
Classifier{okData =
ClassData{prior = 0.0, unseen = -3.332204510175204,
@ -908,13 +951,6 @@ classifiers
[("hourhour", -1.3862943611198906),
("morningtime-of-day (latent)", -1.3862943611198906)],
n = 1}}),
("valentine's day",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("Sonntag",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.1972245773362196,
@ -961,24 +997,22 @@ classifiers
likelihoods = HashMap.fromList [], n = 0}}),
("<day-of-month>(ordinal) <named-month>",
Classifier{okData =
ClassData{prior = -0.2231435513142097,
unseen = -3.4657359027997265,
ClassData{prior = -0.2411620568168881, unseen = -3.367295829986474,
likelihoods =
HashMap.fromList
[("ordinal (digits)Juli", -2.0476928433652555),
("ordinal (digits)September", -2.3353749158170367),
("ordinal (digits)Oktober", -2.740840023925201),
("month", -0.8690378470236094),
("ordinal (digits)Februar", -1.6422277352570913),
("ordinals (first..19th)Marz", -2.740840023925201)],
n = 12},
[("ordinal (digits)Juli", -1.9459101490553135),
("ordinal (digits)September", -2.2335922215070942),
("month", -0.8472978603872037),
("ordinal (digits)Februar", -1.540445040947149),
("ordinals (first..19th)Marz", -2.639057329615259)],
n = 11},
koData =
ClassData{prior = -1.6094379124341003, unseen = -2.639057329615259,
ClassData{prior = -1.540445040947149, unseen = -2.5649493574615367,
likelihoods =
HashMap.fromList
[("ordinal (digits)Juli", -1.466337068793427),
("ordinal (digits)April", -1.8718021769015913),
("month", -1.1786549963416462)],
[("ordinal (digits)Juli", -1.3862943611198906),
("ordinal (digits)April", -1.791759469228055),
("month", -1.0986122886681098)],
n = 3}}),
("numbers prefix with -, negative or minus",
Classifier{okData =
@ -1006,13 +1040,6 @@ classifiers
[("after lunch", -1.252762968495368),
("hour", -1.252762968495368)],
n = 1}}),
("new year's eve",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("tomorrow",
Classifier{okData =
ClassData{prior = -0.13353139262452263,
@ -1022,13 +1049,6 @@ classifiers
ClassData{prior = -2.0794415416798357,
unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1}}),
("Mother's Day",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("<time> after next",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.0794415416798357,
@ -1245,8 +1265,8 @@ classifiers
ClassData{prior = 0.0, unseen = -2.0794415416798357,
likelihoods =
HashMap.fromList
[("a <unit-of-duration>christmas", -1.252762968495368),
("<integer> <unit-of-duration>christmas", -1.252762968495368),
[("<integer> <unit-of-duration>Weihnachten", -1.252762968495368),
("a <unit-of-duration>Weihnachten", -1.252762968495368),
("yearday", -0.8472978603872037)],
n = 2},
koData =
@ -1273,8 +1293,8 @@ classifiers
likelihoods = HashMap.fromList [], n = 0}}),
("Oktober",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.3025850929940455,
likelihoods = HashMap.fromList [("", 0.0)], n = 8},
ClassData{prior = 0.0, unseen = -2.1972245773362196,
likelihoods = HashMap.fromList [("", 0.0)], n = 7},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
@ -1505,15 +1525,14 @@ classifiers
likelihoods = HashMap.fromList [], n = 0}}),
("<day-of-month> (ordinal)",
Classifier{okData =
ClassData{prior = -0.5520685823000397,
unseen = -3.0910424533583156,
ClassData{prior = -0.5753641449035618, unseen = -3.044522437723423,
likelihoods =
HashMap.fromList
[("ordinals (first..19th)", -1.6582280766035324),
("ordinal (digits)", -0.2113090936672069)],
n = 19},
[("ordinals (first..19th)", -1.6094379124341003),
("ordinal (digits)", -0.2231435513142097)],
n = 18},
koData =
ClassData{prior = -0.8574502318512216, unseen = -2.833213344056216,
ClassData{prior = -0.8266785731844679, unseen = -2.833213344056216,
likelihoods =
HashMap.fromList
[("ordinals (first..19th)", -0.8266785731844679),
@ -1627,13 +1646,6 @@ classifiers
("day", -1.791759469228055),
("day (grain)", -1.791759469228055)],
n = 2}}),
("christmas",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.791759469228055,
likelihoods = HashMap.fromList [("", 0.0)], n = 4},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("Tag der Deutschen Einheit",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
@ -1641,13 +1653,6 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("new year's day",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.3862943611198906,
likelihoods = HashMap.fromList [("", 0.0)], n = 2},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("next n <cycle>",
Classifier{okData =
ClassData{prior = 0.0, unseen = -4.143134726391533,
@ -1677,13 +1682,6 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -3.1354942159291497,
likelihoods = HashMap.fromList [], n = 0}}),
("halloween day",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("by the end of <time>",
Classifier{okData =
ClassData{prior = -infinity, unseen = -1.6094379124341003,
@ -1823,9 +1821,9 @@ classifiers
ClassData{prior = -0.6931471805599453, unseen = -1.791759469228055,
likelihoods =
HashMap.fromList
[("dayday", -0.916290731874155),
("ordinals (first..19th)Dienstagchristmas",
-0.916290731874155)],
[("ordinals (first..19th)DienstagWeihnachten",
-0.916290731874155),
("dayday", -0.916290731874155)],
n = 1}}),
("<named-month> <day-of-month> (non ordinal)",
Classifier{okData =
@ -1893,10 +1891,11 @@ classifiers
likelihoods = HashMap.fromList [], n = 0}}),
("ordinal (digits)",
Classifier{okData =
ClassData{prior = -0.2876820724517809, unseen = -2.995732273553991,
likelihoods = HashMap.fromList [("", 0.0)], n = 18},
ClassData{prior = -0.3022808718729337,
unseen = -2.9444389791664407,
likelihoods = HashMap.fromList [("", 0.0)], n = 17},
koData =
ClassData{prior = -1.3862943611198906,
ClassData{prior = -1.3437347467010947,
unseen = -2.0794415416798357,
likelihoods = HashMap.fromList [("", 0.0)], n = 6}}),
("quarter (grain)",
@ -1969,10 +1968,10 @@ classifiers
HashMap.fromList
[("<time-of-day> o'clock", -2.3025850929940455),
("intersect", -2.3025850929940455),
("Weihnachten", -1.8971199848858813),
("day", -1.6094379124341003),
("time-of-day (latent)", -2.3025850929940455),
("hh:mm", -2.3025850929940455), ("hour", -1.8971199848858813),
("christmas", -1.8971199848858813),
("minute", -2.3025850929940455)],
n = 6}}),
("day (grain)",

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

View File

@ -454,13 +454,6 @@ classifiers
("<hour-of-day> and <relative minutes>", -1.791759469228055),
("hour", -1.791759469228055), ("minute", -1.791759469228055)],
n = 4}}),
("season",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.3862943611198906,
likelihoods = HashMap.fromList [("", 0.0)], n = 2},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("minutos (grain)",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.1972245773362196,
@ -620,6 +613,13 @@ classifiers
koData =
ClassData{prior = 0.0, unseen = -2.639057329615259,
likelihoods = HashMap.fromList [("", 0.0)], n = 12}}),
("invierno",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("this <part-of-day>",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.6094379124341003,
@ -701,6 +701,13 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("verano",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("number (20..90)",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.791759469228055,
@ -1243,13 +1250,14 @@ classifiers
likelihoods = HashMap.fromList [], n = 0}}),
("ce <time>",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.772588722239781,
ClassData{prior = 0.0, unseen = -2.833213344056216,
likelihoods =
HashMap.fromList
[("season", -1.6094379124341003), ("Lunes", -2.0149030205422647),
("day", -1.3217558399823195), ("hour", -1.6094379124341003),
("week-end", -1.6094379124341003)],
[("invierno", -2.0794415416798357),
("Lunes", -2.0794415416798357), ("verano", -2.0794415416798357),
("day", -1.3862943611198906), ("hour", -1.6739764335716716),
("week-end", -1.6739764335716716)],
n = 5},
koData =
ClassData{prior = -infinity, unseen = -1.791759469228055,
ClassData{prior = -infinity, unseen = -1.9459101490553135,
likelihoods = HashMap.fromList [], n = 0}})]

View File

@ -21,7 +21,14 @@ import Duckling.Ranking.Types
classifiers :: Classifiers
classifiers
= HashMap.fromList
[("<time> timezone",
[("Kerstmis",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.791759469228055,
likelihoods = HashMap.fromList [("", 0.0)], n = 4},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("<time> timezone",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.3978952727983707,
likelihoods =
@ -65,13 +72,6 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -1.791759469228055,
likelihoods = HashMap.fromList [], n = 0}}),
("Father's Day",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("thursday",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.4849066497880004,
@ -249,6 +249,13 @@ classifiers
[("hh:mmtime-of-day (latent)", -1.0986122886681098),
("minutehour", -1.0986122886681098)],
n = 1}}),
("Halloween",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("between <datetime> and <datetime> (interval)",
Classifier{okData =
ClassData{prior = -0.6931471805599453,
@ -380,268 +387,260 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("May",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("intersect",
Classifier{okData =
ClassData{prior = -0.30818059400012193,
unseen = -6.182084906716631,
ClassData{prior = -0.29944691403136736,
unseen = -6.180016653652572,
likelihoods =
HashMap.fromList
[("last <cycle>sunday", -5.486869473092628),
("hourday", -4.388257184424518),
[("last <cycle>sunday", -5.484796933490655),
("Koningsdagyear", -5.484796933490655),
("hourday", -4.386184644822546),
("on a named-day<day-of-month> (non ordinal) <named-month>",
-5.486869473092628),
-5.484796933490655),
("<day-of-month> (non ordinal) <named-month><time-of-day> o'clock",
-5.486869473092628),
("dayhour", -2.3513752571634776),
("daymonth", -4.570578741218473),
("next <cycle>friday", -5.486869473092628),
("monthyear", -3.8774315606585272),
-5.484796933490655),
("dayhour", -2.3493027175615055),
("daymonth", -4.5685062016165),
("next <cycle>friday", -5.484796933490655),
("monthyear", -3.8753590210565547),
("friday<day-of-month> (non ordinal) <named-month>",
-5.081404364984463),
("yearhour", -5.486869473092628),
("intersect<time-of-day> o'clock", -4.793722292532682),
("tuesdaythis <cycle>", -5.486869473092628),
("christmasyear", -5.486869473092628),
-5.079331825382491),
("yearhour", -5.484796933490655),
("intersect<time-of-day> o'clock", -4.791649752930709),
("tuesdaythis <cycle>", -5.484796933490655),
("on a named-dayfrom <time-of-day> - <time-of-day> (interval)",
-5.081404364984463),
("saturday<time-of-day> o'clock", -5.486869473092628),
-5.079331825382491),
("saturday<time-of-day> o'clock", -5.484796933490655),
("absorption of , after named daydd/mm/yyyy",
-5.081404364984463),
("<time-of-day> am|pmintersect", -5.486869473092628),
("intersect<time> <part-of-day>", -4.793722292532682),
("Marchyear", -5.486869473092628),
("saturdayat <time-of-day>", -5.081404364984463),
("saturday<time> <part-of-day>", -5.081404364984463),
("dd/mm<time-of-day> o'clock", -5.486869473092628),
("on a named-dayat <time-of-day>", -4.793722292532682),
("today<time-of-day> o'clock", -5.486869473092628),
("at <time-of-day>on a named-day", -5.486869473092628),
-5.079331825382491),
("<time-of-day> am|pmintersect", -5.484796933490655),
("intersect<time> <part-of-day>", -4.791649752930709),
("Marchyear", -5.484796933490655),
("saturdayat <time-of-day>", -5.079331825382491),
("saturday<time> <part-of-day>", -5.079331825382491),
("dd/mm<time-of-day> o'clock", -5.484796933490655),
("on a named-dayat <time-of-day>", -4.791649752930709),
("today<time-of-day> o'clock", -5.484796933490655),
("at <time-of-day>on a named-day", -5.484796933490655),
("on a named-daybetween <datetime> and <datetime> (interval)",
-5.486869473092628),
("intersect by ','year", -5.486869473092628),
("on <date><time-of-day> o'clock", -4.793722292532682),
("next <cycle>monday", -5.486869473092628),
-5.484796933490655),
("intersect by ','year", -5.484796933490655),
("on <date><time-of-day> o'clock", -4.791649752930709),
("next <cycle>monday", -5.484796933490655),
("on a named-daybetween <time-of-day> and <time-of-day> (interval)",
-5.486869473092628),
("exactly <time-of-day>tomorrow", -5.486869473092628),
("on a named-day<time> <part-of-day>", -5.081404364984463),
("monthhour", -4.570578741218473),
-5.484796933490655),
("exactly <time-of-day>tomorrow", -5.484796933490655),
("on a named-day<time> <part-of-day>", -5.079331825382491),
("monthhour", -4.5685062016165),
("on <date>between <datetime> and <datetime> (interval)",
-5.486869473092628),
("last <day-of-week> of <time>year", -5.486869473092628),
("on a named-day<time-of-day> o'clock", -5.486869473092628),
("next <cycle>March", -5.486869473092628),
("todayat <time-of-day>", -4.793722292532682),
-5.484796933490655),
("last <day-of-week> of <time>year", -5.484796933490655),
("on a named-day<time-of-day> o'clock", -5.484796933490655),
("next <cycle>March", -5.484796933490655),
("todayat <time-of-day>", -4.791649752930709),
("on <date>between <time-of-day> and <time-of-day> (interval)",
-5.486869473092628),
("on <date>at <time-of-day>", -4.793722292532682),
("thursday<time> timezone", -5.081404364984463),
("on <date>about <time-of-day>", -4.793722292532682),
("dayday", -3.184284380098582),
("on <date><time> <part-of-day>", -4.570578741218473),
("dd/mmabout <time-of-day>", -5.081404364984463),
("<day-of-month> (ordinal)intersect", -5.486869473092628),
("hourhour", -4.388257184424518),
-5.484796933490655),
("on <date>at <time-of-day>", -4.791649752930709),
("thursday<time> timezone", -5.079331825382491),
("on <date>about <time-of-day>", -4.791649752930709),
("dayday", -3.1822118404966093),
("on <date><time> <part-of-day>", -4.5685062016165),
("dd/mmabout <time-of-day>", -5.079331825382491),
("<day-of-month> (ordinal)intersect", -5.484796933490655),
("hourhour", -4.386184644822546),
("saturday<day-of-month> (non ordinal) <named-month>",
-5.486869473092628),
("thursdayat <time-of-day>", -5.486869473092628),
-5.484796933490655),
("thursdayat <time-of-day>", -5.484796933490655),
("today/tomorrow/yesterday afternoon<hour-of-day> <integer> (as relative minutes)",
-5.081404364984463),
-5.079331825382491),
("intersect<day-of-month> (non ordinal) <named-month>",
-4.388257184424518),
("dayyear", -3.4074279314127915),
("tuesdayafter <duration>", -5.486869473092628),
("tomorrow<time-of-day> o'clock", -5.486869473092628),
("<time-of-day> o'clocktomorrow", -4.793722292532682),
-4.386184644822546),
("dayyear", -3.405355391810819),
("tuesdayafter <duration>", -5.484796933490655),
("tomorrow<time-of-day> o'clock", -5.484796933490655),
("<time-of-day> o'clocktomorrow", -4.791649752930709),
("<day-of-month> (non ordinal) <named-month>intersect",
-5.486869473092628),
-5.484796933490655),
("thursdaybetween <time-of-day> and <time-of-day> (interval)",
-5.486869473092628),
-5.484796933490655),
("thursdaybetween <datetime> and <datetime> (interval)",
-5.486869473092628),
-5.484796933490655),
("<day-of-month>(ordinal) <named-month>year",
-5.486869473092628),
("<day-of-month> (ordinal)tuesday", -5.486869473092628),
("hourminute", -4.793722292532682),
("thursdayhh:mm", -5.486869473092628),
("this <cycle>wednesday", -5.486869473092628),
-5.484796933490655),
("<day-of-month> (ordinal)tuesday", -5.484796933490655),
("hourminute", -4.791649752930709),
("thursdayhh:mm", -5.484796933490655),
("this <cycle>wednesday", -5.484796933490655),
("absorption of , after named day<named-month> <day-of-month> (non ordinal)",
-5.486869473092628),
-5.484796933490655),
("<day-of-month> (non ordinal) <named-month><time> <part-of-day>",
-5.081404364984463),
-5.079331825382491),
("on <date><day-of-month> (non ordinal) <named-month>",
-5.486869473092628),
("Aprilabout <time-of-day>", -5.486869473092628),
-5.484796933490655),
("Aprilabout <time-of-day>", -5.484796933490655),
("thursdayfrom <datetime> - <datetime> (interval)",
-5.486869473092628),
("<day-of-month> (ordinal)February", -5.486869473092628),
-5.484796933490655),
("<day-of-month> (ordinal)February", -5.484796933490655),
("thursdayfrom <time-of-day> - <time-of-day> (interval)",
-4.793722292532682),
-4.791649752930709),
("absorption of , after named day<day-of-month> (non ordinal) <named-month>",
-4.570578741218473),
-4.5685062016165),
("today/tomorrow/yesterday evening<time-of-day> o'clock",
-5.486869473092628),
("fridayintersect", -4.793722292532682),
-5.484796933490655),
("fridayintersect", -4.791649752930709),
("on <date><time-of-day> - <time-of-day> (interval)",
-5.486869473092628),
-5.484796933490655),
("on <date><datetime> - <datetime> (interval)",
-5.486869473092628),
("this <cycle>monday", -5.486869473092628),
("April<time-of-day> o'clock", -5.486869473092628),
("next <cycle>wednesday", -5.486869473092628),
("hh:mmintersect", -5.081404364984463),
-5.484796933490655),
("this <cycle>monday", -5.484796933490655),
("April<time-of-day> o'clock", -5.484796933490655),
("next <cycle>wednesday", -5.484796933490655),
("hh:mmintersect", -5.079331825382491),
("thursday<datetime> - <datetime> (interval)",
-5.081404364984463),
("<day-of-month> (ordinal)July", -5.486869473092628),
-5.079331825382491),
("<day-of-month> (ordinal)July", -5.484796933490655),
("thursday<time-of-day> - <time-of-day> (interval)",
-5.081404364984463),
("yearmonth", -5.486869473092628),
("at <time-of-day>intersect", -5.081404364984463),
-5.079331825382491),
("yearmonth", -5.484796933490655),
("at <time-of-day>intersect", -5.079331825382491),
("on <date>from <time-of-day> - <time-of-day> (interval)",
-5.081404364984463),
-5.079331825382491),
("<time> <part-of-day>from <time-of-day> - <time-of-day> (interval)",
-5.486869473092628),
("dayminute", -2.7788192719904172),
-5.484796933490655),
("dayminute", -2.7767467323884447),
("<time> <part-of-day>from <datetime> - <datetime> (interval)",
-5.486869473092628),
("intersectyear", -4.793722292532682),
("<time-of-day> am|pmsaturday", -5.486869473092628),
("<ordinal> <cycle> of <time>year", -5.486869473092628),
("minuteday", -3.289644895756408),
-5.484796933490655),
("intersectyear", -4.791649752930709),
("<time-of-day> am|pmsaturday", -5.484796933490655),
("<ordinal> <cycle> of <time>year", -5.484796933490655),
("minuteday", -3.2875723561544357),
("absorption of , after named dayintersect",
-5.486869473092628),
("Julyyear", -5.486869473092628),
("year<time-of-day> o'clock", -5.486869473092628),
("Octoberyear", -4.388257184424518),
("hh:mmon <date>", -5.081404364984463),
-5.484796933490655),
("Julyyear", -5.484796933490655),
("year<time-of-day> o'clock", -5.484796933490655),
("Octoberyear", -4.386184644822546),
("hh:mmon <date>", -5.079331825382491),
("today/tomorrow/yesterday afternoonat <time-of-day>",
-5.486869473092628),
("dd/mmyear", -5.486869473092628),
("tuesdayin <duration>", -5.486869473092628),
("Septemberyear", -5.081404364984463),
("at <time-of-day>on <date>", -5.081404364984463),
("absorption of , after named daydd/mm", -5.081404364984463),
("intersectintersect", -5.486869473092628),
("hh:mmon a named-day", -5.486869473092628),
("weekday", -4.100575111972737),
("dayweek", -4.793722292532682),
("weekyear", -5.081404364984463),
("on <date>February", -5.486869473092628),
("<month> dd-dd (interval)year", -5.486869473092628),
("Dutch King dayyear", -5.486869473092628),
("tomorrowat <time-of-day>", -5.081404364984463),
("next <cycle>tuesday", -5.486869473092628),
-5.484796933490655),
("dd/mmyear", -5.484796933490655),
("tuesdayin <duration>", -5.484796933490655),
("Septemberyear", -5.079331825382491),
("at <time-of-day>on <date>", -5.079331825382491),
("absorption of , after named daydd/mm", -5.079331825382491),
("intersectintersect", -5.484796933490655),
("hh:mmon a named-day", -5.484796933490655),
("weekday", -4.098502572370764),
("dayweek", -4.791649752930709),
("weekyear", -5.079331825382491),
("on <date>February", -5.484796933490655),
("Kerstmisyear", -5.484796933490655),
("<month> dd-dd (interval)year", -5.484796933490655),
("tomorrowat <time-of-day>", -5.079331825382491),
("next <cycle>tuesday", -5.484796933490655),
("monday<day-of-month> (non ordinal) <named-month>",
-5.486869473092628),
("at <time-of-day>tomorrow", -5.486869473092628),
-5.484796933490655),
("at <time-of-day>tomorrow", -5.484796933490655),
("monday<named-month> <day-of-month> (non ordinal)",
-5.486869473092628),
-5.484796933490655),
("<day-of-month> (non ordinal) <named-month>about <time-of-day>",
-5.486869473092628),
-5.484796933490655),
("on a named-day<time-of-day> - <time-of-day> (interval)",
-5.486869473092628),
-5.484796933490655),
("today/tomorrow/yesterday eveningat <time-of-day>",
-5.081404364984463),
("last <cycle> of <time>year", -5.081404364984463),
-5.079331825382491),
("last <cycle> of <time>year", -5.079331825382491),
("on a named-day<datetime> - <datetime> (interval)",
-5.486869473092628),
-5.484796933490655),
("<day-of-month> (non ordinal) <named-month>year",
-4.570578741218473),
("April<time> <part-of-day>", -5.081404364984463)],
-4.5685062016165),
("April<time> <part-of-day>", -5.079331825382491)],
n = 169},
koData =
ClassData{prior = -1.327205444749884, unseen = -5.5909869805108565,
ClassData{prior = -1.3518081850487211, unseen = -5.572154032177765,
likelihoods =
HashMap.fromList
[("<named-month> <day-of-month> (non ordinal)July",
-4.894101477840304),
("dayhour", -3.5078071167204135),
("on a named-dayafter <time-of-day>", -4.894101477840304),
-4.875197323201151),
("dayhour", -3.488902962081261),
("on a named-dayafter <time-of-day>", -4.875197323201151),
("July<day-of-month> (non ordinal) <named-month>",
-4.894101477840304),
("daymonth", -3.977810745966149),
("monthday", -4.48863636973214),
("monthyear", -3.7954891891721947),
("yearhour", -4.894101477840304),
("intersect<time-of-day> o'clock", -4.894101477840304),
-4.875197323201151),
("daymonth", -3.9589065913269965),
("monthday", -4.4697322150929875),
("monthyear", -3.776585034533042),
("yearhour", -4.875197323201151),
("intersect<time-of-day> o'clock", -4.875197323201151),
("on a named-dayfrom <time-of-day> - <time-of-day> (interval)",
-4.48863636973214),
("<day-of-month> (ordinal)wednesday", -4.48863636973214),
("Julyintersect", -4.894101477840304),
("thursdayafter <time-of-day>", -4.894101477840304),
("intersect<time> <part-of-day>", -4.894101477840304),
("Marchyear", -4.894101477840304),
("<day-of-month> (ordinal)year", -4.894101477840304),
("on a named-dayat <time-of-day>", -4.894101477840304),
-4.4697322150929875),
("<day-of-month> (ordinal)wednesday", -4.4697322150929875),
("Julyintersect", -4.875197323201151),
("thursdayafter <time-of-day>", -4.875197323201151),
("intersect<time> <part-of-day>", -4.875197323201151),
("Marchyear", -4.875197323201151),
("<day-of-month> (ordinal)year", -4.875197323201151),
("on a named-dayat <time-of-day>", -4.875197323201151),
("on a named-daybetween <datetime> and <datetime> (interval)",
-4.894101477840304),
("tuesdayafter <time-of-day>", -4.894101477840304),
-4.875197323201151),
("tuesdayafter <time-of-day>", -4.875197323201151),
("on a named-daybetween <time-of-day> and <time-of-day> (interval)",
-4.894101477840304),
("monthhour", -3.977810745966149),
-4.875197323201151),
("monthhour", -3.9589065913269965),
("on <date>between <datetime> and <datetime> (interval)",
-4.894101477840304),
("todayat <time-of-day>", -4.894101477840304),
-4.875197323201151),
("todayat <time-of-day>", -4.875197323201151),
("on <date>between <time-of-day> and <time-of-day> (interval)",
-4.894101477840304),
("on <date>at <time-of-day>", -4.894101477840304),
("on <date>about <time-of-day>", -4.894101477840304),
("dayday", -3.5078071167204135),
("hourhour", -4.200954297280359),
("thursdayat <time-of-day>", -4.894101477840304),
("dayyear", -4.200954297280359),
-4.875197323201151),
("on <date>at <time-of-day>", -4.875197323201151),
("on <date>about <time-of-day>", -4.875197323201151),
("dayday", -3.776585034533042),
("hourhour", -4.182050142641207),
("thursdayat <time-of-day>", -4.875197323201151),
("dayyear", -4.182050142641207),
("thursdaybetween <time-of-day> and <time-of-day> (interval)",
-4.894101477840304),
-4.875197323201151),
("thursdaybetween <datetime> and <datetime> (interval)",
-4.894101477840304),
("<day-of-month> (ordinal)tuesday", -4.48863636973214),
("thursdayhh:mm", -4.48863636973214),
("Aprilabout <time-of-day>", -4.894101477840304),
-4.875197323201151),
("<day-of-month> (ordinal)tuesday", -4.4697322150929875),
("thursdayhh:mm", -4.4697322150929875),
("Aprilabout <time-of-day>", -4.875197323201151),
("thursdayfrom <datetime> - <datetime> (interval)",
-4.894101477840304),
("Aprilyear", -4.894101477840304),
("<day-of-month> (ordinal)sunday", -4.48863636973214),
-4.875197323201151),
("Aprilyear", -4.875197323201151),
("thursdayfrom <time-of-day> - <time-of-day> (interval)",
-4.200954297280359),
("mondayFebruary", -4.894101477840304),
-4.182050142641207),
("mondayFebruary", -4.875197323201151),
("on <date><time-of-day> - <time-of-day> (interval)",
-4.894101477840304),
("yearuntil <time-of-day>", -4.48863636973214),
-4.875197323201151),
("yearuntil <time-of-day>", -4.4697322150929875),
("thursday<time-of-day> - <time-of-day> (interval)",
-4.48863636973214),
-4.4697322150929875),
("<named-month> <day-of-month> (non ordinal)August",
-4.894101477840304),
-4.875197323201151),
("August<day-of-month> (non ordinal) <named-month>",
-4.894101477840304),
-4.875197323201151),
("on <date>from <time-of-day> - <time-of-day> (interval)",
-4.48863636973214),
("dayminute", -2.409194828052304),
("thursday<part-of-day> of <time>", -4.894101477840304),
("Julyyear", -4.48863636973214),
("Octoberyear", -4.894101477840304),
-4.4697322150929875),
("dayminute", -2.3902906734131513),
("thursday<part-of-day> of <time>", -4.875197323201151),
("Julyyear", -4.4697322150929875),
("Octoberyear", -4.875197323201151),
("today/tomorrow/yesterday afternoonat <time-of-day>",
-4.48863636973214),
("year<time> <part-of-day>", -4.894101477840304),
-4.4697322150929875),
("year<time> <part-of-day>", -4.875197323201151),
("today/tomorrow/yesterday afternoon<time-of-day> o'clock",
-4.894101477840304),
("yearday", -4.48863636973214),
("absorption of , after named dayFebruary", -4.894101477840304),
-4.875197323201151),
("yearday", -4.4697322150929875),
("absorption of , after named dayFebruary", -4.875197323201151),
("<day-of-month> (non ordinal) <named-month>about <time-of-day>",
-4.894101477840304),
("after <time-of-day>year", -4.894101477840304),
-4.875197323201151),
("after <time-of-day>year", -4.875197323201151),
("on a named-day<time-of-day> - <time-of-day> (interval)",
-4.894101477840304),
("on <date>after <time-of-day>", -4.894101477840304),
-4.875197323201151),
("on <date>after <time-of-day>", -4.875197323201151),
("<day-of-month> (non ordinal) <named-month>year",
-4.894101477840304)],
n = 61}}),
-4.875197323201151)],
n = 59}}),
("<ordinal> <cycle> of <time>",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.4849066497880004,
@ -820,27 +819,24 @@ classifiers
likelihoods = HashMap.fromList [], n = 0}}),
("nth <time> of <time>",
Classifier{okData =
ClassData{prior = -0.40546510810816444,
unseen = -3.044522437723423,
ClassData{prior = -0.5596157879354228, unseen = -2.70805020110221,
likelihoods =
HashMap.fromList
[("daymonth", -1.0498221244986778),
("ordinals (first..19th)sundayMay", -2.3025850929940455),
("ordinals (first..19th)tuesdayOctober", -2.3025850929940455),
("ordinals (first..19th)sundayJune", -2.3025850929940455),
[("daymonth", -1.0296194171811581),
("ordinals (first..19th)tuesdayOctober", -1.9459101490553135),
("ordinals (first..19th)wednesdayintersect",
-1.8971199848858813),
-1.540445040947149),
("ordinals (first..19th)tuesdayintersect",
-2.3025850929940455)],
n = 6},
-1.9459101490553135)],
n = 4},
koData =
ClassData{prior = -1.0986122886681098, unseen = -2.70805020110221,
ClassData{prior = -0.8472978603872037,
unseen = -2.5649493574615367,
likelihoods =
HashMap.fromList
[("daymonth", -1.252762968495368),
("ordinals (first..19th)wednesdayOctober", -1.540445040947149),
("ordinals (first..19th)tuesdaySeptember",
-1.9459101490553135)],
[("daymonth", -1.0986122886681098),
("ordinals (first..19th)wednesdayOctober", -1.3862943611198906),
("ordinals (first..19th)tuesdaySeptember", -1.791759469228055)],
n = 3}}),
("the <day-of-month> (non ordinal)",
Classifier{okData =
@ -882,7 +878,7 @@ classifiers
[("hourhour", -0.6931471805599453),
("morningtime-of-day (latent)", -0.6931471805599453)],
n = 1}}),
("valentine's day",
("Vaderdag",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
@ -937,7 +933,7 @@ classifiers
[("integer (numeric)", -9.53101798043249e-2),
("fractional number", -2.3978952727983707)],
n = 20}}),
("new year's eve",
("Nieuwjaarsdag",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.3862943611198906,
likelihoods = HashMap.fromList [("", 0.0)], n = 2},
@ -953,13 +949,6 @@ classifiers
ClassData{prior = -1.3862943611198906,
unseen = -1.3862943611198906,
likelihoods = HashMap.fromList [("", 0.0)], n = 2}}),
("Mother's Day",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("friday",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.3025850929940455,
@ -992,13 +981,6 @@ classifiers
koData =
ClassData{prior = 0.0, unseen = -2.0794415416798357,
likelihoods = HashMap.fromList [("", 0.0)], n = 6}}),
("June",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("afternoon",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.5649493574615367,
@ -1045,8 +1027,8 @@ classifiers
likelihoods = HashMap.fromList [], n = 0}}),
("sunday",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.3978952727983707,
likelihoods = HashMap.fromList [("", 0.0)], n = 9},
ClassData{prior = 0.0, unseen = -2.1972245773362196,
likelihoods = HashMap.fromList [("", 0.0)], n = 7},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
@ -1153,8 +1135,8 @@ classifiers
n = 68}}),
("ordinals (first..19th)",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.5649493574615367,
likelihoods = HashMap.fromList [("", 0.0)], n = 11},
ClassData{prior = 0.0, unseen = -2.3978952727983707,
likelihoods = HashMap.fromList [("", 0.0)], n = 9},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
@ -1176,8 +1158,8 @@ classifiers
ClassData{prior = 0.0, unseen = -1.6094379124341003,
likelihoods =
HashMap.fromList
[("<integer> <unit-of-duration>christmas", -0.6931471805599453),
("yearday", -0.6931471805599453)],
[("yearday", -0.6931471805599453),
("<integer> <unit-of-duration>Kerstmis", -0.6931471805599453)],
n = 1},
koData =
ClassData{prior = -infinity, unseen = -1.0986122886681098,
@ -1292,7 +1274,7 @@ classifiers
likelihoods = HashMap.fromList [], n = 0}}),
("<day-of-month> (ordinal)",
Classifier{okData =
ClassData{prior = -1.3862943611198906,
ClassData{prior = -1.2809338454620642,
unseen = -2.0794415416798357,
likelihoods =
HashMap.fromList
@ -1300,12 +1282,19 @@ classifiers
("ordinal (digits)", -0.5596157879354228)],
n = 5},
koData =
ClassData{prior = -0.2876820724517809, unseen = -2.890371757896165,
ClassData{prior = -0.325422400434628, unseen = -2.772588722239781,
likelihoods =
HashMap.fromList
[("ordinals (first..19th)", -0.5306282510621704),
("ordinal (digits)", -0.8873031950009028)],
n = 15}}),
[("ordinals (first..19th)", -0.6286086594223742),
("ordinal (digits)", -0.7621400520468967)],
n = 13}}),
("Valentijnsdag",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("until <time-of-day>",
Classifier{okData =
ClassData{prior = -0.40546510810816444,
@ -1402,20 +1391,6 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -1.9459101490553135,
likelihoods = HashMap.fromList [], n = 0}}),
("christmas",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.791759469228055,
likelihoods = HashMap.fromList [("", 0.0)], n = 4},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("new year's day",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.3862943611198906,
likelihoods = HashMap.fromList [("", 0.0)], n = 2},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("Sinterklaas",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.6094379124341003,
@ -1467,7 +1442,7 @@ classifiers
[("<integer> <unit-of-duration>", -0.6931471805599453),
("hour", -0.6931471805599453)],
n = 1}}),
("halloween day",
("Koningsdag",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
@ -1624,8 +1599,8 @@ classifiers
ClassData{prior = -0.6931471805599453, unseen = -1.791759469228055,
likelihoods =
HashMap.fromList
[("ordinals (first..19th)tuesdaychristmas", -0.916290731874155),
("dayday", -0.916290731874155)],
[("dayday", -0.916290731874155),
("ordinals (first..19th)tuesdayKerstmis", -0.916290731874155)],
n = 1}}),
("<named-month> <day-of-month> (non ordinal)",
Classifier{okData =
@ -1741,6 +1716,13 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("Oudjaar",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.3862943611198906,
likelihoods = HashMap.fromList [("", 0.0)], n = 2},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("after <time-of-day>",
Classifier{okData =
ClassData{prior = -0.587786664902119, unseen = -2.995732273553991,
@ -1755,9 +1737,9 @@ classifiers
ClassData{prior = -0.8109302162163288, unseen = -2.890371757896165,
likelihoods =
HashMap.fromList
[("intersect", -2.1400661634962708),
[("Kerstmis", -1.7346010553881064),
("intersect", -2.1400661634962708),
("day", -1.4469189829363254), ("hh:mm", -2.1400661634962708),
("christmas", -1.7346010553881064),
("minute", -2.1400661634962708)],
n = 4}}),
("day (grain)",
@ -1767,6 +1749,13 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("Moederdag",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("<month> dd-dd (interval)",
Classifier{okData =
ClassData{prior = 0.0, unseen = -2.639057329615259,
@ -1778,13 +1767,6 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -1.3862943611198906,
likelihoods = HashMap.fromList [], n = 0}}),
("Dutch King day",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,
likelihoods = HashMap.fromList [("", 0.0)], n = 1},
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("summer",
Classifier{okData =
ClassData{prior = 0.0, unseen = -1.0986122886681098,

View File

@ -193,25 +193,6 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("last tuesday, last july",
Classifier{okData =
ClassData{prior = -1.2039728043259361,
unseen = -3.4965075614664802,
likelihoods =
HashMap.fromList
[("day", -0.9007865453381898), ("Sunday", -1.3862943611198906),
("Tuesday", -1.6739764335716716)],
n = 12},
koData =
ClassData{prior = -0.35667494393873245,
unseen = -4.174387269895637,
likelihoods =
HashMap.fromList
[("<integer> (latent time-of-day)", -1.0233888674305223),
("month (numeric with month symbol)", -2.367123614131617),
("February", -3.4657359027997265),
("hour", -1.0233888674305223), ("month", -2.2129729343043585)],
n = 28}}),
("next <cycle>",
Classifier{okData =
ClassData{prior = -0.7472144018302211,

View File

@ -193,25 +193,6 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("last tuesday, last july",
Classifier{okData =
ClassData{prior = -1.2039728043259361,
unseen = -3.4965075614664802,
likelihoods =
HashMap.fromList
[("day", -0.9007865453381898), ("Sunday", -1.3862943611198906),
("Tuesday", -1.6739764335716716)],
n = 12},
koData =
ClassData{prior = -0.35667494393873245,
unseen = -4.174387269895637,
likelihoods =
HashMap.fromList
[("<integer> (latent time-of-day)", -1.0233888674305223),
("month (numeric with month symbol)", -2.367123614131617),
("February", -3.4657359027997265),
("hour", -1.0233888674305223), ("month", -2.2129729343043585)],
n = 28}}),
("next <cycle>",
Classifier{okData =
ClassData{prior = -0.7472144018302211,

View File

@ -193,25 +193,6 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("last tuesday, last july",
Classifier{okData =
ClassData{prior = -1.2039728043259361,
unseen = -3.4965075614664802,
likelihoods =
HashMap.fromList
[("day", -0.9007865453381898), ("Sunday", -1.3862943611198906),
("Tuesday", -1.6739764335716716)],
n = 12},
koData =
ClassData{prior = -0.35667494393873245,
unseen = -4.174387269895637,
likelihoods =
HashMap.fromList
[("<integer> (latent time-of-day)", -1.0233888674305223),
("month (numeric with month symbol)", -2.367123614131617),
("February", -3.4657359027997265),
("hour", -1.0233888674305223), ("month", -2.2129729343043585)],
n = 28}}),
("next <cycle>",
Classifier{okData =
ClassData{prior = -0.7472144018302211,

View File

@ -193,25 +193,6 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("last tuesday, last july",
Classifier{okData =
ClassData{prior = -1.2039728043259361,
unseen = -3.4965075614664802,
likelihoods =
HashMap.fromList
[("day", -0.9007865453381898), ("Sunday", -1.3862943611198906),
("Tuesday", -1.6739764335716716)],
n = 12},
koData =
ClassData{prior = -0.35667494393873245,
unseen = -4.174387269895637,
likelihoods =
HashMap.fromList
[("<integer> (latent time-of-day)", -1.0233888674305223),
("month (numeric with month symbol)", -2.367123614131617),
("February", -3.4657359027997265),
("hour", -1.0233888674305223), ("month", -2.2129729343043585)],
n = 28}}),
("next <cycle>",
Classifier{okData =
ClassData{prior = -0.7472144018302211,

View File

@ -186,25 +186,6 @@ classifiers
koData =
ClassData{prior = -infinity, unseen = -0.6931471805599453,
likelihoods = HashMap.fromList [], n = 0}}),
("last tuesday, last july",
Classifier{okData =
ClassData{prior = -1.2039728043259361,
unseen = -3.4965075614664802,
likelihoods =
HashMap.fromList
[("day", -0.9007865453381898), ("Sunday", -1.3862943611198906),
("Tuesday", -1.6739764335716716)],
n = 12},
koData =
ClassData{prior = -0.35667494393873245,
unseen = -4.174387269895637,
likelihoods =
HashMap.fromList
[("<integer> (latent time-of-day)", -1.0233888674305223),
("month (numeric with month symbol)", -2.367123614131617),
("February", -3.4657359027997265),
("hour", -1.0233888674305223), ("month", -2.2129729343043585)],
n = 28}}),
("next <cycle>",
Classifier{okData =
ClassData{prior = -0.7472144018302211,

View File

@ -79,7 +79,6 @@ negativeCorpus = (testContext, testOptions, examples)
, "we are separated"
, "25"
, "this is the one"
, "in 61"
, "this one"
, "this past one"
, "at single"
@ -110,6 +109,9 @@ latentCorpus = (testContext, testOptions {withLatent = True}, xs)
, examples (datetime (2013, 2, 12, 7, 0, 0) Hour)
[ "7"
]
, examples (datetime (1954, 1, 1, 0, 0, 0) Year)
[ "1954"
]
, examples (datetimeInterval
((2013, 2, 12, 4, 0, 0), (2013, 2, 12, 12, 0, 0)) Hour)
[ "morning"
@ -126,9 +128,6 @@ latentCorpus = (testContext, testOptions {withLatent = True}, xs)
((2013, 2, 12, 18, 0, 0), (2013, 2, 13, 0, 0, 0)) Hour)
[ "night"
]
, examples (datetime (1954, 1, 1, 0, 0, 0) Year)
[ "1954"
]
]
allExamples :: [Example]
@ -145,6 +144,9 @@ allExamples = concat
, examples (datetime (2013, 2, 1, 0, 0, 0) Day)
[ "2/2013"
]
, examples (datetime (2014, 1, 1, 0, 0, 0) Year)
[ "in 2014"
]
, examples (datetime (2013, 2, 11, 0, 0, 0) Day)
[ "yesterday"
]

View File

@ -47,7 +47,7 @@ ruleIntersect = Rule
]
, prod = \tokens -> case tokens of
(Token Time td1:Token Time td2:_) ->
Token Time <$> intersect td1 td2 -- . notLatent
Token Time . notLatent <$> intersect td1 td2
_ -> Nothing
}
@ -89,15 +89,15 @@ ruleAbsorbOnADOW = Rule
_ -> Nothing
}
ruleAbsorbInMonth :: Rule
ruleAbsorbInMonth = Rule
{ name = "in <named-month>"
ruleAbsorbInMonthYear :: Rule
ruleAbsorbInMonthYear = Rule
{ name = "in <named-month>|year"
, pattern =
[ regex "in"
, Predicate isAMonth
, Predicate $ or . sequence [isAMonth, isGrainOfTime TG.Year]
]
, prod = \tokens -> case tokens of
(_:token:_) -> Just token
(_:Token Time td:_) -> tt $ notLatent td
_ -> Nothing
}
@ -1992,7 +1992,7 @@ rules =
, ruleIntersectOf
, ruleAbsorbOnTime
, ruleAbsorbOnADOW
, ruleAbsorbInMonth
, ruleAbsorbInMonthYear
, ruleAbsorbCommaTOD
, ruleNextDOW
, ruleNextTime

View File

@ -52,7 +52,7 @@ tests = testGroup "EN Tests"
, intersectTests
, rangeTests
, localeTests
, latentTests
, makeCorpusTest [This Time] latentCorpus
]
localeTests :: TestTree
@ -175,6 +175,3 @@ rangeTests = testCase "Range Test" $
, ("one ninety nine a m", Range 11 19) -- ruleMilitarySpelledOutAMPM2
, ("thirteen fifty nine a m", Range 15 23) -- ruleMilitarySpelledOutAMPM
]
latentTests :: TestTree
latentTests = makeCorpusTest [This Time] latentCorpus