Recent Advances in MULTIFLEX, a Morphological Generator of - PowerPoint PPT Presentation

Recent Advances in MULTIFLEX, a Morphological Generator of Multi-Word Units Agata Savary November 17, 2008

Morfeusz/Multiflex Platform ◮ Aim: description of morphology and variation of compounds ◮ Information on the language level: classes, categories and values of the IPIPAN tagset ◮ Generating inflected forms of single words with Morfeusz ◮ Combining (by a graph ) inflected forms of single words to create inflected forms of compounds ◮ Unification and value inheritance for a compact description

Example 1: Maria Skłodowska-Curie Maria Skłodowska-Curie, Marii Skłodowskiej-Curie, . . . Skłodowska-Curie, Skłodowskiej-Curie, . . . Maria Skłodowska, Marii Skłodowskiej, . . . Maria Curie, Marii Curie, . . . Skłodowska, Skłodowskiej, . . .

Annotation of components and inflection graph Maria Skłodowska - Curie $1 $2 $3 $4 $5 lemma: Maria lemma: Skłodowska class: subst class: subst homonym: 0 homonym: 0 Nb: sg Nb: sg Case : nom Case : nom Gen: f Gen: f <$1:Case=$c> <$2> <$3:Case=$c> <$4> <$5> <Gen=$1.Gen;Nb=$1.Nb;Case=$c>

Annotated forms Maria Skłodowska-Curie Maria Skłodowska-Curie:subst:sg: nom :f Marii Skłodowskiej-Curie Maria Skłodowska-Curie:subst:sg: gen :f Marii Skłodowskiej-Curie Maria Skłodowska-Curie:subst:sg: dat :f Skłodowską-Curie Maria Skłodowska-Curie:subst:sg: inst :f Skłodowskiej Maria Skłodowska-Curie:subst:sg: loc :f

Example 2: ulica Marii Skłodowskiej-Curie ulica Marii Skłodowskiej-Curie, ulicy Marii Skłodowskiej-Curie, . . . ; ulica Marii Skłodowskiej, ulicy Marii Skłodowskiej, . . . ; ulica Marii Curie, ulicy Marii Curie, . . . ; ulica Skłodowskiej-Curie, ulicy Skłodowskiej-Curie, . . . ; ulica Skłodowskiej, ulicy Skłodowskiej, . . . ; Marii Skłodowskiej-Curie ; Marii Skłodowskiej ; Marii Curie ; Skłodowskiej-Curie ; Skłodowskiej

Flat description of variants: ulica Marii Skłodowskiej-Curie ulica Marii Skłodowskiej - Curie $1 $2 $3 $4 $5 $6 $7 lemma: ulica class: subst homonym: 0 Nb: sg Case : nom Gen: f <$5> <$6> <$7> <$1:Case=$c> <$2> <$3> <$4> <Gen=$1.Gen;Nb=$1.Nb;Case=$c>

New solution : embedded description of ulica Marii Skłodowskiej-Curie ulica Marii Skłodowskiej-Curie $1 $2 $3 lemma: ulica lemma: Maria Skłodowska-Curie class: subst class: subst homonym: 0 homonym: 0 Nb: sg Nb: sg Case : nom Case : gen Gen: f Gen: f <$1:Case=$c> <$2> <$3:Case=gen> <Gen=$1.Gen;Nb=$1.Nb;Case=$c>

Morphology of numerals in the IPIPAN tagset Two kinds of “traditional” numerals : ◮ Cardinal numerals ( num ) - a class on its own ◮ have a fixed number ◮ inflect for case, gender, and accomodability ◮ complex morpho-syntactic behaviour ◮ Ordinal numerals - behave morphologically as adjectives ◮ have a fixed number ◮ inflect for case, gender, and degree Problem: how to describe multi-word numerals ? - their number is infinite - their are spelled with letters or digits (dwudziesty vs. 20.) + their vocabulary is small + their rules of creation are very regular

Ordinal numerals - what we wish ◮ A canonical form for each numeral: { dziesięcio-tysięczny, 10-cio-tysięczny, 10-tysięczny, 10000., 10 000. } ⇒ 10000 ◮ A complete annotation: 10000:adj:sg:m1:nom , etc. (?) ◮ Morphological analysis: dziesięcio-tysięczny ⇒ 10000:adj:sg:m1.m2.m3:nom ◮ Morphological generation: 10000:adj:sg:m1:nom ⇒ { dziesięcio-tysięczny, 10-cio-tysięczny, 10-tysięczny, 10000., 10 000. }

Graph-based description: ordinal numerals 0-99 (simplified) Info inside the boxes: textual variants Info under the boxes: canonical forms

Sub-graph: LiczPorz1-9

Perspective: using graphs for “inflection” of compound numerals in MULTIFLEX ulica 11 Listopada 1918 roku $1 $2 $3 $4 $5 $6 $7 $8 $9 lemma: 11 lemma: 1918 lemma: ulica class: adj class: adj class: subst homonym: 0 homonym: 0 homonym: 0 Nb: sg Nb: sg Nb: sg Case: gen Case: gen Case : nom Gen: m3 Gen: m3 Gen: f Deg: pos Deg: pos

“Inflected forms” to be generated ulica 11 Listopada 1918 roku ulica 11 Listopada 1918 11 Listopada 1918 roku ulica Jednastego Listopada 1918 roku ulica 11-ego Listopada 1918 roku ulica 11 Listopada Tysiąc Dziewięćset Osiemnastego roku ? ulica 11 Listopada Osiemnastego roku etc.

One step further ulica 11 Listopada 1918 roku $1 $2 $3 lemma: ulica lemma: 11.11.1918 class: subst class: subst? homonym: 0 homonym: 0 Nb: sg Nb: sg Case : nom Case: gen Gen: f Gen: m3 Graphs describe possible dates and their variants: 11.11.1918 11/11/1918 11 listopada 1918 11 listopada osiemnastego roku jedenasty listopada osiemnastego roku etc.

Recent Advances in MULTIFLEX, a Morphological Generator of - PowerPoint PPT Presentation

Recent Advances in MULTIFLEX, a Morphological Generator of Multi-Word Units Agata Savary November 17, 2008 Morfeusz/Multiflex Platform Aim: description of morphology and variation of compounds Information on the language level: classes,

Recent Advances in Photonic Recent Advances in Photonic effect employing IP- based distributed

Recent Advances in Biomolecular NMR Lucia Banci CERM University of Florence Recent Advances

Recent Advances in Biomolecular NMR Lucia Banci CERM University of Florence Recent Advances

Confl flict and Development: Recent Advances and Future Agendas Professor Patricia Justino

Recent Advances Aly Khawaja Outline STAR- CCM+: a complete simulation workflow Emphasis on

Recent advances in Mandelbrot martingales theory Julien Barral, Universit e Paris Nord

Recent Advances in Adap.ve Sampling and Reconstruc.on for Monte

Recent Advances in Adaptive Sampling and Reconstruction for Monte

Recent Advances in Two-loop Superstrings Eric DHoker Institut des Hautes Etudes Scientifiques,

Recent Advances in Adversarial Machine Learning Nicholas Carlini Google Research Recent

Recent advances on the acceleration of first-order methods in convex optimization . Juan

Some Recent Advances in Nonnegative Matrix Factorization and their Applications to Hyperspectral

Thank you for joining us. The program will commence momentarily. Recent Advances in Medical

Recent Theoretical Advances in Sparse Approximation Joel A. Tropp

Thank you for joining us. The program will commence momentarily. Recent Advances in Medical

Accuracy and Stability: recent advances in C.A.G.D. J.M. Pe na* *University of Zaragoza, Spain

Thank you for joining us. The program will commence momentarily. Recent Advances in Medical

From Declarative Signatures to Misuse IDS 4th International Symposium on Recent Advances in

Becoming Human Purpose: to re-evaluate recent advances in in th the sciences, and develop a

MULTIFLEX - a Formalism and a Tool for the Computational Morphology of Multi-Word Units Agata

Recent Advances in Simulation for Marine Applications Milovan Peri CD-adapco, Nrnberg Office

International conference on Recent Advances in Materials Science and Biophysics Department of

Recent advances in fermentation technology for value-addition to cotton stalks and cottonseed Dr.

Recent advances in the development of ice cloud bulk scattering and absorption models for use with