from dependency structures to lfg representations
play

From dependency structures to LFG representations Dag Haug Seminar - PowerPoint PPT Presentation

Mar 09, 2024 •187 likes •1.65k views

The corpus Conversion LFG101 F-structures C-structure Conclusions From dependency structures to LFG representations Dag Haug Seminar in computational linguistics April 18 Dag Haug dg2lfg April 18 1 / 57 The corpus Conversion LFG101

  1. The corpus Conversion LFG101 F-structures C-structure Conclusions The syntactic annotation scheme: dependency grammar Information about syntactic relations and word order stored separately Reliance on overt elements Inherent problems of: (asyndetic) coordination, structure sharing Dependency grammar with LFG adjustments Limited set of empty nodes (for asyndetic coordination and ellipsis) Dag Haug dg2lfg April 18 8 / 57

  2. The corpus Conversion LFG101 F-structures C-structure Conclusions The syntactic annotation scheme: dependency grammar Information about syntactic relations and word order stored separately Reliance on overt elements Inherent problems of: (asyndetic) coordination, structure sharing Dependency grammar with LFG adjustments Limited set of empty nodes (for asyndetic coordination and ellipsis) Secondary dependencies (for structure sharing, incl. control/raising) Dag Haug dg2lfg April 18 8 / 57

  3. The corpus Conversion LFG101 F-structures C-structure Conclusions The syntactic annotation scheme: dependency grammar Information about syntactic relations and word order stored separately Reliance on overt elements Inherent problems of: (asyndetic) coordination, structure sharing Dependency grammar with LFG adjustments Limited set of empty nodes (for asyndetic coordination and ellipsis) Secondary dependencies (for structure sharing, incl. control/raising) More granular syntactic relations than usual Dag Haug dg2lfg April 18 8 / 57

  4. The corpus Conversion LFG101 F-structures C-structure Conclusions Syntactic relations Label Label Function Function PRED Predicate XADV Free predicative SUB Subject XOBJ Open complement OBJ Object Aux Auxiliary OBL Oblique XOBJ Open complement clause AG Agent COMP Complement clause ADV Adverbial PART Partitive ATR Attribute PARPRED Parenthetical APOS Apposition VOC Vocative NARG Nominal argument example Dag Haug dg2lfg April 18 9 / 57

  5. The corpus Conversion LFG101 F-structures C-structure Conclusions Empty nodes Null conjunctions for asyndetic parataxis Null verbs for null copulas and elided verbs Dag Haug dg2lfg April 18 10 / 57

  6. The corpus Conversion LFG101 F-structures C-structure Conclusions Eliminability of empty nodes Dag Haug dg2lfg April 18 11 / 57

  7. The corpus Conversion LFG101 F-structures C-structure Conclusions Eliminability of empty nodes Dag Haug dg2lfg April 18 11 / 57

  8. The corpus Conversion LFG101 F-structures C-structure Conclusions Human processing of which the Belgians inhabit one, the Aquitani V another,C those who are called Celts in their own language – C Gauls V in our – V the third. Dag Haug dg2lfg April 18 12 / 57

  9. The corpus Conversion LFG101 F-structures C-structure Conclusions Human processing of which the Belgians inhabit one, the Aquitani V another,C those who are called Celts in their own language – C Gauls V in our – V the third. Dag Haug dg2lfg April 18 12 / 57

  10. The corpus Conversion LFG101 F-structures C-structure Conclusions Structure sharing Subject control: Example Object control: Example Various other possibilities Could also be encoded in the label but typically not with the same precision Dag Haug dg2lfg April 18 13 / 57

  11. The corpus Conversion LFG101 F-structures C-structure Conclusions Projectivity language source nonprojective projective Latin Gallic War 1887 22717 Letters to Atticus 2006 20416 Vulgate 4217 92186 Per. Aeth. 1279 14890 Greek Herodotus 6606 56175 NT 4377 103418 OCS Zographensis 36 1034 Suprasliensis 416 7780 Marianus 1828 47731 Gothic NT 1886 46884 Armenian NT 1231 59556 Koriwn 48 1556 Dag Haug dg2lfg April 18 14 / 57

  12. The corpus Conversion LFG101 F-structures C-structure Conclusions Token alignments The translations of the NT have been aligned with the Greek original Dag Haug dg2lfg April 18 15 / 57

  13. The corpus Conversion LFG101 F-structures C-structure Conclusions Token alignments The translations of the NT have been aligned with the Greek original A ‘dictionary’ based on likelihood of occurring in the same bible verse Dag Haug dg2lfg April 18 15 / 57

  14. The corpus Conversion LFG101 F-structures C-structure Conclusions Token alignments The translations of the NT have been aligned with the Greek original A ‘dictionary’ based on likelihood of occurring in the same bible verse Information from the annotation: syntax, morphology, word order Dag Haug dg2lfg April 18 15 / 57

  15. The corpus Conversion LFG101 F-structures C-structure Conclusions Token alignments The translations of the NT have been aligned with the Greek original A ‘dictionary’ based on likelihood of occurring in the same bible verse Information from the annotation: syntax, morphology, word order Manual correction of the Slavic indicates very good results (and a very literal translation) Precision Recall F-score 95.27% 92.97% 94.11% Dag Haug dg2lfg April 18 15 / 57

  16. The corpus Conversion LFG101 F-structures C-structure Conclusions Givenness Givenness tags based on which context the hearer uses to establish reference Discourse (anaphora) → OLD Dag Haug dg2lfg April 18 16 / 57

  17. The corpus Conversion LFG101 F-structures C-structure Conclusions Givenness Givenness tags based on which context the hearer uses to establish reference Discourse (anaphora) → OLD Situation (deixis) → ACC-sit Dag Haug dg2lfg April 18 16 / 57

  18. The corpus Conversion LFG101 F-structures C-structure Conclusions Givenness Givenness tags based on which context the hearer uses to establish reference Discourse (anaphora) → OLD Situation (deixis) → ACC-sit Scenarios (inferences) → ACC-inf Dag Haug dg2lfg April 18 16 / 57

  19. The corpus Conversion LFG101 F-structures C-structure Conclusions Givenness Givenness tags based on which context the hearer uses to establish reference Discourse (anaphora) → OLD Situation (deixis) → ACC-sit Scenarios (inferences) → ACC-inf Encyclopedic knowledge → ACC-gen Dag Haug dg2lfg April 18 16 / 57

  20. The corpus Conversion LFG101 F-structures C-structure Conclusions Givenness Givenness tags based on which context the hearer uses to establish reference Discourse (anaphora) → OLD Situation (deixis) → ACC-sit Scenarios (inferences) → ACC-inf Encyclopedic knowledge → ACC-gen No context (no extra-NP information) → NEW Dag Haug dg2lfg April 18 16 / 57

  21. The corpus Conversion LFG101 F-structures C-structure Conclusions Givenness Givenness tags based on which context the hearer uses to establish reference Discourse (anaphora) → OLD Situation (deixis) → ACC-sit Scenarios (inferences) → ACC-inf Encyclopedic knowledge → ACC-gen No context (no extra-NP information) → NEW example Dag Haug dg2lfg April 18 16 / 57

  22. The corpus Conversion LFG101 F-structures C-structure Conclusions Modal subordination Luke 5:39 Und niemand ist, der vom alten trinkt und wolle bald den neuen; denn er spricht: Der alte ist milder. The subject and the old and the new wine are embedded under subordination Should be inaccessible (Karttunen, COLING 69) but they aren’t We ignore recursive embeddings and use a special tagset for all embedded referents Dag Haug dg2lfg April 18 17 / 57

  23. The corpus Conversion LFG101 F-structures C-structure Conclusions Tagset for embedded referents nonspec (but quant for quantification) nonspec inf nonspec old Dag Haug dg2lfg April 18 18 / 57

  24. The corpus Conversion LFG101 F-structures C-structure Conclusions Tagset for embedded referents nonspec (but quant for quantification) nonspec inf nonspec old No counterparts to acc-gen or acc-sit as these belong in the main DRS by definition Dag Haug dg2lfg April 18 18 / 57

  25. The corpus Conversion LFG101 F-structures C-structure Conclusions Interannotator agreement Towards the end of the NT tagging projects, kappa values were around 0.8 (after long periods of weekly meetings) Dag Haug dg2lfg April 18 19 / 57

  26. The corpus Conversion LFG101 F-structures C-structure Conclusions Interannotator agreement Towards the end of the NT tagging projects, kappa values were around 0.8 (after long periods of weekly meetings) New project: Caesar’s Gallic War Dag Haug dg2lfg April 18 19 / 57

  27. The corpus Conversion LFG101 F-structures C-structure Conclusions Interannotator agreement Towards the end of the NT tagging projects, kappa values were around 0.8 (after long periods of weekly meetings) New project: Caesar’s Gallic War Supervised tagging of 8 chapters (ca. 400 taggables) Dag Haug dg2lfg April 18 19 / 57

  28. The corpus Conversion LFG101 F-structures C-structure Conclusions Interannotator agreement Towards the end of the NT tagging projects, kappa values were around 0.8 (after long periods of weekly meetings) New project: Caesar’s Gallic War Supervised tagging of 8 chapters (ca. 400 taggables) Unsupervised tagging of 5 chapters (ca. 250 taggables) κ = 0.66 counting divergences in taggables κ = 0.75 on tags set by both annotators Dag Haug dg2lfg April 18 19 / 57

  29. The corpus Conversion LFG101 F-structures C-structure Conclusions Interannotator agreement Towards the end of the NT tagging projects, kappa values were around 0.8 (after long periods of weekly meetings) New project: Caesar’s Gallic War Supervised tagging of 8 chapters (ca. 400 taggables) Unsupervised tagging of 5 chapters (ca. 250 taggables) κ = 0.66 counting divergences in taggables κ = 0.75 on tags set by both annotators Decent; but much potential for more agreement, especially in taggables Dag Haug dg2lfg April 18 19 / 57

  30. The corpus Conversion LFG101 F-structures C-structure Conclusions Size of IS corpus Tag Freq old 34430 old inact 1395 acc gen 3755 acc inf 2634 edge type freq acc sit 883 coreference 36650 new 5768 bridging 2847 kind 1178 total 39497 non spec 4485 non spec inf 408 non spec old 1799 quant 2021 total 58756 Dag Haug dg2lfg April 18 20 / 57

  31. The corpus Conversion LFG101 F-structures C-structure Conclusions Storing linguistic analyses Theory-neutrality → data for larger audiences Dag Haug dg2lfg April 18 21 / 57

  32. The corpus Conversion LFG101 F-structures C-structure Conclusions Storing linguistic analyses Theory-neutrality → data for larger audiences widening gulf between corpus linguistics and linguistic theory Dag Haug dg2lfg April 18 21 / 57

  33. The corpus Conversion LFG101 F-structures C-structure Conclusions Storing linguistic analyses Theory-neutrality → data for larger audiences widening gulf between corpus linguistics and linguistic theory DG corpora (Prague, PROIEL) → DG not really in use as a linguistic theory Dag Haug dg2lfg April 18 21 / 57

  34. The corpus Conversion LFG101 F-structures C-structure Conclusions Storing linguistic analyses Theory-neutrality → data for larger audiences widening gulf between corpus linguistics and linguistic theory DG corpora (Prague, PROIEL) → DG not really in use as a linguistic theory PS corpora (Penn, NEGRA) typically use flatter tree structures than anyone believes in Dag Haug dg2lfg April 18 21 / 57

  35. The corpus Conversion LFG101 F-structures C-structure Conclusions Storing linguistic analyses Theory-neutrality → data for larger audiences widening gulf between corpus linguistics and linguistic theory DG corpora (Prague, PROIEL) → DG not really in use as a linguistic theory PS corpora (Penn, NEGRA) typically use flatter tree structures than anyone believes in On the other hand, LFG and HPSG corpora can be hard to use for people who do not share the theoretical assumptions of these theories Dag Haug dg2lfg April 18 21 / 57

  36. The corpus Conversion LFG101 F-structures C-structure Conclusions Our take Principles 1 Encode no more structure than is common to all frameworks Dag Haug dg2lfg April 18 22 / 57

  37. The corpus Conversion LFG101 F-structures C-structure Conclusions Our take Principles 1 Encode no more structure than is common to all frameworks 2 Enoded structure could be seen as derived/secondary in some frameworks Dag Haug dg2lfg April 18 22 / 57

  38. The corpus Conversion LFG101 F-structures C-structure Conclusions Our take Principles 1 Encode no more structure than is common to all frameworks 2 Enoded structure could be seen as derived/secondary in some frameworks 3 Encode enough structure to allow reconstruction of theoretically motived structures Dag Haug dg2lfg April 18 22 / 57

  39. The corpus Conversion LFG101 F-structures C-structure Conclusions Our take Principles 1 Encode no more structure than is common to all frameworks 2 Enoded structure could be seen as derived/secondary in some frameworks 3 Encode enough structure to allow reconstruction of theoretically motived structures In the ideal situation, the information in the annotation can be (monotonically) expanded to structures conforming to a particular theory by adding information from the assumptions of that theory Dag Haug dg2lfg April 18 22 / 57

  40. The corpus Conversion LFG101 F-structures C-structure Conclusions The ideal situation The added assumptions will typically be about phrase structure, such as various versions of X ′ theory Dag Haug dg2lfg April 18 23 / 57

  41. The corpus Conversion LFG101 F-structures C-structure Conclusions The ideal situation The added assumptions will typically be about phrase structure, such as various versions of X ′ theory Given information about what the subject is, it will be possible to create a structure where the subject has a specific position if the theory requires that (unless the data contradict the theory) Dag Haug dg2lfg April 18 23 / 57

  42. The corpus Conversion LFG101 F-structures C-structure Conclusions The ideal situation The added assumptions will typically be about phrase structure, such as various versions of X ′ theory Given information about what the subject is, it will be possible to create a structure where the subject has a specific position if the theory requires that (unless the data contradict the theory) Useful for hypothesis testing Dag Haug dg2lfg April 18 23 / 57

  43. The corpus Conversion LFG101 F-structures C-structure Conclusions Basic principles Modular: several levels of grammatical description connected by projections (functions) Dag Haug dg2lfg April 18 24 / 57

  44. The corpus Conversion LFG101 F-structures C-structure Conclusions Basic principles Modular: several levels of grammatical description connected by projections (functions) The c-structure is a tree structure described by a CFG Dag Haug dg2lfg April 18 24 / 57

  45. The corpus Conversion LFG101 F-structures C-structure Conclusions Basic principles Modular: several levels of grammatical description connected by projections (functions) The c-structure is a tree structure described by a CFG The f-structure is a set of ordered attribute-value pairs Dag Haug dg2lfg April 18 24 / 57

  46. The corpus Conversion LFG101 F-structures C-structure Conclusions Basic principles Modular: several levels of grammatical description connected by projections (functions) The c-structure is a tree structure described by a CFG The f-structure is a set of ordered attribute-value pairs the attribute is a grammatical function or feature and the value is a symbol a semantic form an f-structure a set of f-structures (for adjuncts) Dag Haug dg2lfg April 18 24 / 57

  47. The corpus Conversion LFG101 F-structures C-structure Conclusions Basic principles Modular: several levels of grammatical description connected by projections (functions) The c-structure is a tree structure described by a CFG The f-structure is a set of ordered attribute-value pairs the attribute is a grammatical function or feature and the value is a symbol a semantic form an f-structure a set of f-structures (for adjuncts) Lexical items and CFG rules can contribute f-descriptions Dag Haug dg2lfg April 18 24 / 57

  48. The corpus Conversion LFG101 F-structures C-structure Conclusions Basic principles Modular: several levels of grammatical description connected by projections (functions) The c-structure is a tree structure described by a CFG The f-structure is a set of ordered attribute-value pairs the attribute is a grammatical function or feature and the value is a symbol a semantic form an f-structure a set of f-structures (for adjuncts) Lexical items and CFG rules can contribute f-descriptions Lexical-functional languages ∈ context-sensitive languages Dag Haug dg2lfg April 18 24 / 57

  49. The corpus Conversion LFG101 F-structures C-structure Conclusions Configurational encoding IP 1 1 subj = 2 2 = 3 1 = 4 4 = 5 ↑ subj = ↓ ↑ = ↓ 4 obj = 6 I ′ NP 2 4 6 = 7 ↑ = ↓ N 3 ↑ = ↓ ↑ obj = ↓ I 5 NP 6 Max pred =‘Max’ pushed ↑ = ↓ pred ‘push � subj, obj � ’ N 7 tense = past Fred pred =‘Fred’ Dag Haug dg2lfg April 18 25 / 57

  50. The corpus Conversion LFG101 F-structures C-structure Conclusions Configurational encoding � �   IP 1 pred ‘push subj, obj ’   � �  subj pred ‘Max’    2,3 ↑ subj = ↓ ↑ = ↓     � � I ′ NP 2  pred ‘Fred’  4 obj   6,7   ↑ = ↓ tense past N 3 ↑ = ↓ ↑ obj = ↓ 1,4,5 I 5 NP 6 Max pred =‘Max’ pushed ↑ = ↓ pred ‘push � subj, obj � ’ N 7 tense = past Fred pred =‘Fred’ Dag Haug dg2lfg April 18 25 / 57

  51. The corpus Conversion LFG101 F-structures C-structure Conclusions Structure sharing IP 1 ↑ subj = ↓ ↑ = ↓ I ′ NP 2 4  � �  pred ‘seem xcomp , subj’ ↑ = ↓   � �   subj pred ‘Max’   N 3      � �  ↑ = ↓ ↑ xcomp = ↓  pred ‘push subj, obj ’  Max       I 5 VP  xcomp subj —  pred =‘Max’          � �  obj pred ‘Fred’   seemed     V ↑ obj = ↓ pred ‘seem � xcomp � , subj ’ tense past NP 6 ↑ subj = ↑ xcomp subj push tense = past ‘push � subj, obj � ’ ↑ = ↓ N 7 Fred pred =‘Fred’ Dag Haug dg2lfg April 18 26 / 57

  52. The corpus Conversion LFG101 F-structures C-structure Conclusions Non-configurational encoding S 1 1 gf = 2 2 = 3 ∃ f . f subj = 3 ↑ gf = ↓ ↑ = ↓ ↑ gf = ↓ NP 2 I 4 NP 5 1 = 4 ↑ = ↓ trusit ↑ = ↓ N 3 pred ‘push � subj, obj � ’ N 6 4 gf = 5 tense = past Maximilianus Fredericum 5 = 6 subj ↑ obj ↑ pred =‘Max.’ pred =‘Fred.’ ∃ f . f obj = 6 Dag Haug dg2lfg April 18 27 / 57

  53. The corpus Conversion LFG101 F-structures C-structure Conclusions Non-configurational encoding � �   S 1 pred ‘push subj, obj ’   � �   subj pred ‘Max.’   2,3     � � ↑ gf = ↓ ↑ = ↓ ↑ gf = ↓  obj pred ‘Fred.’    NP 2 I 4 NP 5 6,7   tense past ↑ = ↓ trusit ↑ = ↓ 1,4,5 N 3 pred ‘push � subj, obj � ’ N 6 tense = past Maximilianus Fredericum subj ↑ obj ↑ pred =‘Max.’ pred =‘Fred.’ Dag Haug dg2lfg April 18 27 / 57

  54. The corpus Conversion LFG101 F-structures C-structure Conclusions Non-projectivity A mock Latin example Maximilianus bonum trusit Fredericum Maximilian. nom good. acc pushed Frederick. acc Dag Haug dg2lfg April 18 28 / 57

  55. The corpus Conversion LFG101 F-structures C-structure Conclusions Non-projectivity S 1 1 gf = 2 2 = 3 ∃ f . f subj = 3 1 gf = 4 ↑ gf = ↓ ↑ gf = ↓ ↑ = ↓ ↑ gf = ↓ NP 2 NP 4 I 8 NP 9 4 = 5 6 ∈ 5 adj ↑ = ↓ ↑ = ↓ trusit ↑ = ↓ 7 = 6 N ′ N 3 pred ‘push � subj, obj � ’ N 10 5 7 case = acc tense = past 5 case = acc Maximilianus ↑ adj ∈ ↓ Fredericum 1 = 8 subj ↑ AdjP 6 obj ↑ pred =‘Max.’ pred =‘Fred.’ 1 gf = 9 ↑ = ↓ case = acc 9 = 10 Adj 7 5 case = acc ∃ f . f obj = 10 bonum pred ‘good’ case = acc ( adj ∈ ↑ ) case = acc Dag Haug dg2lfg April 18 29 / 57

  56. The corpus Conversion LFG101 F-structures C-structure Conclusions Non-projectivity S 1  � �  ↑ gf = ↓ ↑ gf = ↓ ↑ = ↓ ↑ gf = ↓ pred ‘push subj, obj ’ NP 2 NP 4 I 8 NP 9   � �   subj pred ‘Max.’    2,3  ↑ = ↓ ↑ = ↓ trusit ↑ = ↓      pred ‘Fred.’  N ′ N 3 pred ‘push � subj, obj � ’ N 10   5   tense = past  case acc        Maximilianus ↑ adj ∈ ↓ Fredericum obj      � �   pred ‘good’      subj ↑ AdjP 6 obj ↑   adj    case acc    pred =‘Max.’ pred =‘Fred.’     6,7   ↑ = ↓ 4,5,9,10 case = acc   Adj 7 tense past 1,8 bonum pred ‘good’ case = acc ( adj ∈ ↑ ) case = acc Dag Haug dg2lfg April 18 30 / 57

  57. The corpus Conversion LFG101 F-structures C-structure Conclusions Relationship to DG F-structures and DGs both encode labelled syntactic dependencies Dag Haug dg2lfg April 18 31 / 57

  58. The corpus Conversion LFG101 F-structures C-structure Conclusions Relationship to DG F-structures and DGs both encode labelled syntactic dependencies Two major differences LFG’s structure sharing runs against DG’s unique head principle In DG, every word introduces depth in the graph, whereas multiple words can contribute to the same F-structure (without nesting) Dag Haug dg2lfg April 18 31 / 57

  59. The corpus Conversion LFG101 F-structures C-structure Conclusions Relationship to DG F-structures and DGs both encode labelled syntactic dependencies Two major differences LFG’s structure sharing runs against DG’s unique head principle In DG, every word introduces depth in the graph, whereas multiple words can contribute to the same F-structure (without nesting) We have already given up the unique head principle in our DG Dag Haug dg2lfg April 18 31 / 57

  60. The corpus Conversion LFG101 F-structures C-structure Conclusions Relationship to DG F-structures and DGs both encode labelled syntactic dependencies Two major differences LFG’s structure sharing runs against DG’s unique head principle In DG, every word introduces depth in the graph, whereas multiple words can contribute to the same F-structure (without nesting) We have already given up the unique head principle in our DG The words that do not introduce separate layers of f-structures are typically function words, so they can be identified from the labels Dag Haug dg2lfg April 18 31 / 57

  61. The corpus Conversion LFG101 F-structures C-structure Conclusions Label mapping Function Label LFG Function Label LFG Adverbial adv adj Oblique obl obj θ / obl Agent ag obl AG Parenthetical parpred — Apposition apos adj Partitive part adj Attribute atr adj Predicate pred — Auxiliary aux — Subject sub subj Complement comp comp Vocative voc — Argument of noun narg ≈ obl Free predicative xadv xadj Object obj obj Open complement xobj xcomp Dag Haug dg2lfg April 18 32 / 57

  62. The corpus Conversion LFG101 F-structures C-structure Conclusions A simple example root Each node maps to an attribute-value matrix with pred morphological features and a semantic form amat obj   pred ’pulcher’ case acc puellam     gend fem atr pulchram Dag Haug dg2lfg April 18 33 / 57

  63. The corpus Conversion LFG101 F-structures C-structure Conclusions A simple example root The relations are translated to attributes with the dependents’ pred AVM as value amat       pred ’pulcher’   obj        adj case acc       puellam    gend fem     atr pulchram Dag Haug dg2lfg April 18 34 / 57

  64. The corpus Conversion LFG101 F-structures C-structure Conclusions A simple example root We do this for all nodes in the structure pred   pred ’puella’ amat case acc   obj   gend fem puellam atr pulchram Dag Haug dg2lfg April 18 35 / 57

  65. The corpus Conversion LFG101 F-structures C-structure Conclusions A simple example root The AVMs of the head and the relation+dependent are unified pred amat   pred ’puella’ obj case acc     puellam  gend fem           pred ’pulcher’  atr         adj case acc       pulchram     gend fem   Dag Haug dg2lfg April 18 36 / 57

  66. The corpus Conversion LFG101 F-structures C-structure Conclusions A simple example root The process terminates with the main verb pred NB pred � = pred amat � � pred ’amare � subj, obj � ’ obj puellam atr pulchram Dag Haug dg2lfg April 18 37 / 57

  67. The corpus Conversion LFG101 F-structures C-structure Conclusions A simple example The final result root pred   ’amare � subj, obj � ’ pred amat     pred ’puella’     obj case acc            gend fem  puellam      obj           pred ’pulcher’  atr               case acc adj       pulchram          gend fem      Dag Haug dg2lfg April 18 38 / 57

  68. The corpus Conversion LFG101 F-structures C-structure Conclusions Structure sharing 1 root Conjunct participles pred challenge the unique head principle dixerunt There are two candidate sub heads: the main verb and et the participle subject atv apos sub sub conversi Gaius Aristarchus comites atr Pauli Dag Haug dg2lfg April 18 39 / 57

  69. The corpus Conversion LFG101 F-structures C-structure Conclusions Structure sharing 2 root pred dixerunt With secondary edges we xadv sub can represent both xsub conversi et dependencies sub apos sub Gaius Aristarchus comites atr Pauli Dag Haug dg2lfg April 18 40 / 57

  70. The corpus Conversion LFG101 F-structures C-structure Conclusions F-structure representation   pred ’dico � sub, obl, comp � ’ subj . . .         � � ’convertor � sub � ’  pred      xadj   subj . . .   Dag Haug dg2lfg April 18 41 / 57

  71. The corpus Conversion LFG101 F-structures C-structure Conclusions Features in coordination   num pl     et        pred ’comes � obl � ’              adj      � �      sub sub apos  obl pred ’Paulus’                                pred ’Gaius’     Gaius Aristarchus comites           num sg       atr      � �        adj —                 Pauli     pred ’Aristarchus’                 num sg     The adjunct is a distributive feature Non-distributive features are computed from the set members Number, gender and person are such features Dag Haug dg2lfg April 18 42 / 57

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

Mapping between Dependency Structures and Compositional Semantic Representations LREC 2010 Max

Mapping between Dependency Structures and Compositional Semantic Representations LREC 2010 Max

Intro Source Target Mapping Evaluation Conclusion Mapping between Dependency Structures and Compositional Semantic Representations LREC 2010 Max Jakob, Mark eta Lopatkov a, Valia Kordoni UFAL at Charles University in Prague,

174 views • 14 slides
Dependency Structures and Locales Category Theory Octoberfest, JHU, 26 October 2019 Gershom

Dependency Structures and Locales Category Theory Octoberfest, JHU, 26 October 2019 Gershom

Dependency Structures Distributive Lattices Locales Versioning Free Distributive Lattices Dependency Problems Conclusion Dependency Structures and Locales Category Theory Octoberfest, JHU, 26 October 2019 Gershom Bazerman Awake Security

479 views • 34 slides
Structures Tao Lei Yu Xin, Yuan Zhang, Regina Barzilay, Tommi Jaakkola CSAIL, MIT 1 Dependency

Structures Tao Lei Yu Xin, Yuan Zhang, Regina Barzilay, Tommi Jaakkola CSAIL, MIT 1 Dependency

Low-Rank Tensors for Scoring Dependency Structures Tao Lei Yu Xin, Yuan Zhang, Regina Barzilay, Tommi Jaakkola CSAIL, MIT 1 Dependency Parsing ROOT I ate cake with a fork today PRON VB NN IN DT NN NN Dependency parsing as

588 views • 36 slides
Learning Dependency Structures for Weak Supervision Models Fred Sala , Paroma Varma, Ann He, Alex

Learning Dependency Structures for Weak Supervision Models Fred Sala , Paroma Varma, Ann He, Alex

Learning Dependency Structures for Weak Supervision Models 6:30-9:00 PM, Pacific Ballroom #119 Learning Dependency Structures for Weak Supervision Models Fred Sala , Paroma Varma, Ann He, Alex Ratner, Chris R Learning Dependency Structures for

375 views • 15 slides
Learning Dependency-Based Compositional Semantics Semantic Representations for Textual Inference

Learning Dependency-Based Compositional Semantics Semantic Representations for Textual Inference

Learning Dependency-Based Compositional Semantics Semantic Representations for Textual Inference Workshop Mar. 10, 2012 Percy Liang Google/Stanford joint work with Michael Jordan and Dan Klein Motivating Problem: Question Answering 2

1.74k views • 142 slides
Minimalist Grammars Formalisme en Dependency Structures Sjoerd Dost Myrthe Tielman Logical

Minimalist Grammars Formalisme en Dependency Structures Sjoerd Dost Myrthe Tielman Logical

The Minimalist Program Minimalist Grammars Merge Onion Dependency Structures Block Degree Nestedness Minimalist Grammars Formalisme en Dependency Structures Sjoerd Dost Myrthe Tielman Logical Methods in NLP 03/05/12 Dost, Tielman

812 views • 48 slides
Dependency Parsing Data structures and algorithms for Computational Linguistics III ar

Dependency Parsing Data structures and algorithms for Computational Linguistics III ar

Dependency Parsing Data structures and algorithms for Computational Linguistics III ar ltekin ccoltekin@sfs.uni-tuebingen.de University of Tbingen Seminar fr Sprachwissenschaft Winter Semester 20192020 Introduction

1.45k views • 59 slides
Dependency Grammars Data structures and algorithms for Computational Linguistics III ar

Dependency Grammars Data structures and algorithms for Computational Linguistics III ar

Dependency Grammars Data structures and algorithms for Computational Linguistics III ar ltekin ccoltekin@sfs.uni-tuebingen.de University of Tbingen Seminar fr Sprachwissenschaft Winter Semester 20192020 Where were we?

546 views • 41 slides
Connection connection = dependency without the governor-dependent hierarchy We will not

Connection connection = dependency without the governor-dependent hierarchy We will not

Interpreting and defining connections in dependency structures Sylvain Kahane (Modyco, Universit Paris Nanterre) August 28, 2019, Paris, Depling, Syntaxfest Connection connection = dependency without the governor-dependent hierarchy

504 views • 32 slides
DP-GP-LVM A Bayesian Non-Parametric Model for Learning Multivariate Dependency Structures Andrew

DP-GP-LVM A Bayesian Non-Parametric Model for Learning Multivariate Dependency Structures Andrew

DP-GP-LVM A Bayesian Non-Parametric Model for Learning Multivariate Dependency Structures Andrew R. Lawrence 1 Carl Henrik Ek 2 Neill D. F. Campbell 1 1 University of Bath, UK 2 University of Bristol, UK Proceedings of the 36 th International

418 views • 20 slides
Model structures on the category of complexes of quiver representations Payam Bahiraei (IPM) (A

Model structures on the category of complexes of quiver representations Payam Bahiraei (IPM) (A

Model structures and Hovey pairs Homotopy category of model category Model structure on C (Rep( Q , G )) Some descriptions of D ( Q ) Morphism categories References Model structures on the category of complexes of quiver representations Payam

902 views • 79 slides
Transforming Dependency Structures to LTAG Derivation Trees 13th International Workshop on Tree

Transforming Dependency Structures to LTAG Derivation Trees 13th International Workshop on Tree

Transforming Dependency Structures to LTAG Derivation Trees 13th International Workshop on Tree Adjoining Grammars and Related Formalisms (TAG+13) Caio Corro Joseph Le Roux September 1, 2017 Laboratoire Informatique de Paris Nord (LIPN),

815 views • 64 slides
Dependency Grammars Topological Dependency Trees: A Constraint-based Account of Linear

Dependency Grammars Topological Dependency Trees: A Constraint-based Account of Linear

Dependency Grammars Topological Dependency Trees: A Constraint-based Account of Linear Precedence Extensible Dependency Grammar: A New Methodology Sibel Ciddi NPFL106 - Linguistics 2013 Summer Framework Immediate dependency

656 views • 18 slides
Quadratic Poisson structures on representations and double brackets Vladimir Rubtsov, ITEP,

Quadratic Poisson structures on representations and double brackets Vladimir Rubtsov, ITEP,

Quadratic Poisson structures on representations and double brackets Vladimir Rubtsov, ITEP, Moscow and LAREMA, Universit e dAngers, (joint work with A.Odesskii and V. Sokolov , Theor. Math. Phys. 2012, 171, 442447 , arXiv:1208.2935v1 and

409 views • 39 slides
CSE 326: Data Structures Graph representations Graphs Topological Sort Topological

CSE 326: Data Structures Graph representations Graphs Topological Sort Topological

Agenda Basic graph terminology CSE 326: Data Structures Graph representations Graphs Topological Sort Topological sort Reference: Weiss, Ch. 9 Hal Perkins Spring 2007 Lectures 22-23 2 Some Applications: Graph ADT?

332 views • 6 slides
Paris and Stanford at EPE 2017: Downstream Evaluation of Graph-based Dependency

Paris and Stanford at EPE 2017: Downstream Evaluation of Graph-based Dependency

Paris and Stanford at EPE 2017: Downstream Evaluation of Graph-based Dependency Representations Sebastian Schuster , ric Villemonte de la Clergerie, Marie Candito, Benot Sagot, Christopher Manning, and Djam Seddah Stanford

971 views • 49 slides
In p ro T K in A c tio n : O p e n -S o u rc e S o ftw a re fo r B u ild in g G e rm a n -S p

In p ro T K in A c tio n : O p e n -S o u rc e S o ftw a re fo r B u ild in g G e rm a n -S p

In p ro T K in A c tio n : O p e n -S o u rc e S o ftw a re fo r B u ild in g G e rm a n -S p e a k in g In c re m e n ta l S D S s T im o B a u m a n n , O k k o B u , D a v id S c h la n g e n tim o @ lin g .u n i- p o ts d a m .d e

549 views • 33 slides
How Virtual Machinery Can Bridge The Explanatory Gap In Natural and Artificial Systems.

How Virtual Machinery Can Bridge The Explanatory Gap In Natural and Artificial Systems.

Seminar in School of Biosciences, University of Birmingham 11 May 2010 Talk at SAB2010, Clos-Luc` e, Amboise, France, 28th Aug 2010 How Virtual Machinery Can Bridge The Explanatory Gap In Natural and Artificial Systems. Presented at

1.02k views • 86 slides
Grammar Implementation with Lexicalized Tree Adjoining Grammars and Frame Semantics Syntactic

Grammar Implementation with Lexicalized Tree Adjoining Grammars and Frame Semantics Syntactic

Grammar Implementation with Lexicalized Tree Adjoining Grammars and Frame Semantics Syntactic analyses Laura Kallmeyer, Timm Lichte, Rainer Osswald & Simon Petitjean University of Dsseldorf DGfS CL Fall School, September 12, 2017 SFB 991

1.37k views • 135 slides
Dependency Parsing and Feature-based Parsing Ling 571 Deep Processing Techniques for

Dependency Parsing and Feature-based Parsing Ling 571 Deep Processing Techniques for

Dependency Parsing and Feature-based Parsing Ling 571 Deep Processing Techniques for NLP October 21, 2019 Shane Steinert-Threlkeld 1 Announcements Thanks for the feedback! HW3: mean 92 Handling ungrammaticality:

1.3k views • 70 slides
Grammar Formalisms Head-Driven Phrase Structure Grammar (HPSG) Laura Kallmeyer, Timm Lichte,

Grammar Formalisms Head-Driven Phrase Structure Grammar (HPSG) Laura Kallmeyer, Timm Lichte,

Grammar Formalisms Head-Driven Phrase Structure Grammar (HPSG) Laura Kallmeyer, Timm Lichte, Wolfgang Maier Universit at T ubingen 27.06.2007 HPSG 1 Outline Feature Structure Basics The building blocks of HPSG Analyzing raising and

596 views • 36 slides
What can we learn from implementing Optimality Theory? Tams Bir ELTE Etvs Lornd

What can we learn from implementing Optimality Theory? Tams Bir ELTE Etvs Lornd

Tools for OT Formal OT Gestalt What can we learn from implementing Optimality Theory? Tams Bir ELTE Etvs Lornd University PTA workshop @ GLOW 41, Budapest, April 14, 2018 Tams Bir What can we learn from implementing OT? 1

513 views • 24 slides
The SmartKom Multimodal Corpus Data Collection and EndtoEnd Evaluation Nicole Beringer

The SmartKom Multimodal Corpus Data Collection and EndtoEnd Evaluation Nicole Beringer

The SmartKom Multimodal Corpus Data Collection and EndtoEnd Evaluation Nicole Beringer Institut fr Phonetik und Sprachliche Kommunikation LMU Mnchen The SmartKom Multimodal Corpus Data Collection and EndtoEnd

399 views • 37 slides
Fast Algorithms for Permutation Pattern Detection William Kuszmaul Stanford University Duluth

Fast Algorithms for Permutation Pattern Detection William Kuszmaul Stanford University Duluth

Fast Algorithms for Permutation Pattern Detection William Kuszmaul Stanford University Duluth REU Supervised by Joe Gallian June 26, 2017 P ATTERNS WITH NO A DJACENCY C ONSTRAINTS Permutation contains Pattern 1 8 3 6 4 5 7 2 3 1 2

833 views • 27 slides

More recommend