gf-core.git/src/runtime/c/pgf/scanner.c, branch optimize Unnamed repository; edit this file 'description' to name the repository. https://git.adelon.net/gf-core.git/atom?h=optimize 2020-07-11T07:52:43+00:00 revert to using unconditional probabilities in the different lookup functions 2020-07-11T07:52:43+00:00 krangelov kr.angelov@gmail.com 2020-07-11T07:52:43+00:00 urn:sha1:77a2630ed94265da8f56648f61deeacaf92960b3 split unknown words by spaces 2020-05-16T06:32:00+00:00 krangelov kr.angelov@gmail.com 2020-05-16T06:32:00+00:00 urn:sha1:9a1f982b1403e80650bd8a2abdaf68d293c31b2a lookupCohorts now detects and reports unknown words. Also: 2020-05-14T13:03:30+00:00 krangelov kr.angelov@gmail.com 2020-05-14T13:03:30+00:00 urn:sha1:62bc78380e69af2de3253130204fc45bac00f3f0 - added added two filtering functions: filterLongest and filterBest - updated the PGF service to work with the new API lookup_morpho, lookup_cohorts and fullform_lexicon now report the usual conditional probability. This turns out to be more useful for part of speech tagging 2020-03-05T11:04:42+00:00 krangelov kr.angelov@gmail.com 2020-03-05T11:04:42+00:00 urn:sha1:88f76ef6710180e4351652d7e84566a86def274a fully supported case-insensitive parsing/lookup 2019-06-30T06:48:23+00:00 krangelov kr.angelov@gmail.com 2019-06-30T06:48:23+00:00 urn:sha1:32379a8d1118838e8f3487e1c54ab6eee813e7a5 the parser now ensures that all word senses are in the chart 2019-06-25T10:58:28+00:00 krangelov kr.angelov@gmail.com 2019-06-25T10:58:28+00:00 urn:sha1:b56591c6b6b3315bdb97164dca574c43292f4d98 filter out empty cohorts 2019-05-28T10:42:59+00:00 krangelov kr.angelov@gmail.com 2019-05-28T10:42:59+00:00 urn:sha1:ac039ec74fad8983e8826c5357f1a40d73a0970a API for scanning for cohorts in an arbitrary text 2019-05-28T10:26:00+00:00 krangelov kr.angelov@gmail.com 2019-05-28T10:26:00+00:00 urn:sha1:9f0ea19a1ca7eba160746bdc91c55f7ae4c0b4fc