[17:35:32] okay. progress. i've got "Paris containing church". how do i make "church" into plural "churches"? [17:35:43] woo! [17:35:55] which fun did you use? [17:36:18] if you go from a VP "contain churches" to Adv, you can just make the NP inside the VP plural [17:42:28] mkAgg : CN -> NP; [17:42:29] mkAgg class = [17:42:29] mkNP a_Det [17:42:29] (mkCN (mkCN (PE.mkA "Aggregate") (PE.mkN "Entity")) [17:42:29] (ExtendEng.GerundAdv (mkVP DictEng.comprise_V2 (mkNP class)))); [17:42:43] the Board of Directors of a Corporation is an Aggregate Entity comprising Director . [17:42:48] i want that to say: ... Directors [17:43:40] yes, then you need mkNP : Det -> CN -> NP [17:43:46] where Det is thePl_Det [17:44:12] mkNP thePl_Det class [17:44:25] but i don't want it to say "comprising the Directors", i want it to say "comprising Directors". [17:44:32] i saw that there is the idea of a "noPrep" [17:44:35] sorry yes [17:44:41] just a sec [17:44:45] maybe i want a noPl_Det [17:45:13] stupid to not have aPl_Det [17:45:17] I thought it existed [17:45:38] I mean in the RGL beyond API the tree would just be DetQuant IndefArt NumPl [17:46:09] so you need Quant -> Num -> CN -> NP [17:46:30] with a_Quant and pluralNum [17:46:47] that translates into no article in english [17:46:53] cool, well, (mkNP aPl_Det CN) works [17:46:59] haha okay [17:47:02] thank you [17:47:10] I wonder why it's not in the API [17:47:14] it is [17:47:20] i just didn't realize it was what i wanted [17:47:40] literally, aPl_Det would be "a things" [17:47:42] oh yes it totally is [17:48:37] okay, i am getting closer to having a thing that converts an ontology/type-system/class-hierarchy to natural language [17:48:45] DMN and OWL will be my first sources [17:49:48] yay! [17:50:40] i wonder if i can convert XML into something that would be easy for GF to parse [17:51:25] mm, it needs to have spaces if we want to parse it with haskell runtime [17:51:47] if i'm coming from Haskell i can just do the XML parsing in Haskell directly [17:51:53] yeah [17:51:57] and then drive the PGF from there [17:52:12] maybe it is time to bite the bullet and learn to wrangle XML in Haskell [17:52:15] parsing with GF is often a bad idea [17:52:44] do you want to head back to the flat btw? we're not really concentrating on the talk :-D [17:52:54] where's alexis? [17:53:35] i thikn she is in the industry session, she will finish at 6:30 [17:53:41] okay [17:53:51] right then we can wait for her! [17:54:24] i am finding it useful to occasionally see what the PhDs are working on. TIL SWRL syntax is very very close to Flora-2 [17:54:27] i wonder if that is deliberate [17:54:46] okay, so you are actually concentrating ^^ [17:55:21] i am not concentrating, i am just browsing [17:55:34] ah, Ben Grosof was involved in both SWRL and Flora-2, of course, now it makes sense [17:56:25] i'd like to debrief about your thoughts on MIREL and JURIX at some point, possibly over a glass of sangria [17:56:28] is sangria vegan? [18:07:26] if the wine they use is vegan! but I usually don't bother that much, it's not like the parts of animals used in clarifying wine are the main purpose of growing those animals :-P [18:07:55] and often people don't know if wines are vegan (and it's the first time they hear wine can be non-vegan) and that's just too much hassle [18:08:04] so for my purposes, yeah it's vegan enough :-P [18:10:50] (also the quantities used are tiny, and I consume wine really rarely, so I'm probably eating more insects accidentally than consuming animal products by my wine drinking)