1 / 6

North Saami Verb Profiling

North Saami Verb Profiling. Jussi Ylikoski , Laura A. Janda, Ciprian Gerstenberger , Hanne Eckhoff. Size and Source of database. Note: disambiguation is a major problem a given form can easily have over 10 interpretations About 80,000 words , comprised of :

kemp
Télécharger la présentation

North Saami Verb Profiling

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. North Saami Verb Profiling Jussi Ylikoski, Laura A. Janda, CiprianGerstenberger, Hanne Eckhoff

  2. Size and Source of database • Note: disambiguation is a major problem • a given form caneasily have over 10 interpretations • About 80,000 words, comprisedof: • North Saami “Gold” corpus, disambiguated by hand, about 30,000 words • Johan Turi’s (1910) Muitalus sámiid birra (An account of the Saami). Automatically disambiguated, plus correction of verb forms, about 50,000 words

  3. Protocol for North Saami Derived verbs: Treat all derived verbs as separate lexemes. This includes verbs derived in -goahtit “start X-ing” and also passive verbs (roughly parallel to Russian -ся verbs). Auxiliary verbs: Ignore/remove all forms of “ii” (negation verb) as aux. Ignore/remove all forms of “leat” as aux. Include all other auxiliary verbs as verbs on equal footing with main verbs. [Assumption: We have the following relevant types of constructions: • “ii” aux + negation forms of main verbs • “leat” aux + non-finite forms of main verbs • other aux + non-finite forms of main verbs For 1. and 2. we will take only the forms of the main verbs, but for 3. we will take both the aux and the main verb.] Threshold: Retain/focus on only lemmas that have 20+ forms in the combined corpus.

  4. Form of data COLUMN 1: PASSPORT Gold vs. Turi COLUMN 2: FORM_ID Unique line # and word ID COLUMN 3: FORM The actual form of the verb found (boahtit, boahtime, gorrojuvvon, jáhkkigoahtit) COLUMN 3: LEMMA The infinitive form of the verb (boahtit, gorrojuvvot, jáhkkigoahtit) COLUMN 5: ANALYSIS Full parse of form COLUMN 5: SUBPARADIGM Ind.Prs, Ind.Prt, Imprt, Inf, PrfPrc, etc. COLUMN 6: CONTEXT The entire sentence that the example form is found in.

  5. Example from dataset PASSPORT;FORM_ID;FORM;LEMMA;ANALYSIS;SUBPARADIGM;CONTEXT Gold;g_3_1;Sávan;sávvat;V.TV.Ind.Prs.Sg1;Ind.Prs;"Sávan didjiidelihku , vástidiiEvo Morales . " Translation: I wishyouluck, answeredEvoMorales.

  6. Results • Almost 20,000 verb forms • About 120 verbs cross thethreshold (>20 forms) • Partiallyannnotated for • auxiliary verbs • semanticclassesfollowing Russian National Corpus

More Related