Coling poster

Coling poster

• 1.
  English Proposition Bank Problem:Which frames may target language verbs evoke? buy.01 Roles: A0: buyer A1: thing bought A2: seller A3: price paid A4: benefactive like.01 Roles: A0: liker A1: object of affection give.01 Roles: A0: giver A1: thing given A2: entity given to Alan Akbik, Xinyu Guan and Yunyao Li IBM Research – Almaden, Yale University Multilingual Aliasing for Auto-Generating Proposition Banks How can we enable Universal Semantic Role Labeling across new languages? Generating High Quality Proposition Banks for Multilingual Semantic Role Labeling. Alan Akbik, Laura Chiticariu, Marina Danilevsky, Yunyao Li, Shivakumar Vaithyanathan and Huaiyu Zhu. ACL 2015. Towards Semi-Automatic Generation of Proposition Banks for Low-Resource Languages. Alan Akbik and Yunyao Li. EMNLP 2016. Chinese Universal Proposition Bank French Universal Proposition Bank German Universal Proposition Bank The manA0 boughtBUY.01 a catA1 yesterdayAM-TMP for 100 dollarsA3. HeA0 reallyAM-MNR likesLIKE.01 the catA1. The engineerA0 boughtBUY.01 a carA1 from a dealership in San JoseA2. SheA0 gaveGIVE.01 the carA1 to her daughterA2 as a presentAM-PRD. Her daughterA0 howeverAM-DIS does notAM-NEG likeLIKE.01 the carA1. Frame lexicon Annotated text 买.01 Source: [buy.01] Roles: A0: buyer A1: thing bought A2: seller A3: price paid A4: benefactive 喝.01 Source: [drink.01] Roles: A0: drinker A1: drink 我A0 也 会 为 你A4 买BUY.01 嫁妆A1 。我A0 甚至 可以 帮 你买BUY.01 一 个 冰淇淋 甜筒A1 。我A0 从来不AM-NEG 喝DRINK.01 瓶装 啤酒A1 。 drehen.01 Source: [turn.01] Roles: A0: turner A1: thing turning Am: direction kaufen.01 Source: [buy.01] Roles: A0: buyer A1: thing bought A2: seller A3: price paid A4: benefactive Der Millionär A0 kaufte BUY.01 sich eine Jacht A1 für 1 millionen Euro A3. Nach dem Kauf AM-TMP drehte TURN.01 er A0 sich zur Jacht A0 und begann, einen Film A0 zu drehen FILM.01. arriver.01 Source: [arrive.01] Roles: A1: 'comer' A2: extent A3: start point A4: destination acheter.01 Source: [buy.01] Roles: A0: buyer A1: thing bought A2: seller A3: price paid A4: benefactive Quand j’A1 arrive ARRIVE.01 en FranceA4, jeA0 vais m’acheter BUY.01 une baguetteA1. Et après celaAM-TMP, jeA0 vais revenirRETURN.01 aux États-Unis. Original drehen. 04 SHOOT.03 record on film A0: videographer A1: subject filmed A2: medium drehen. 01 TURN.01 rotation A0: turner A1: thing turning Am: direction drehen. 02 TURN.02 transformation A0: causer of transformation A1: thing changing A2: end state A3: start state record on film A0: recorder A1: thing filmed drehen. 03 FILM.01 incorrect frame, filtered Merged synonyms, merged Verb: drehen Filtered drehen. 01 TURN.01 rotation A0: turner A1: thing turning Am: direction record on film A0: recorder A1: thing filmed drehen. 03 FILM.01 drehen. 01 TURN.01 rotation A0: turner A1: thing turning Am: direction drehen. 03 FILM.01 SHOOT.03 record on film A0: recorder A1: thing filmed record on film A0: videographer A1: subject filmed A2: medium drehen. 04 SHOOT.03 kosten.01 TASTE.01 use one’s tastebuds A0: taster A1: food Kosten Sie erst den Cocktail! Taste you first the drink! (You haven’t tasted your drink.) Example: kosten.02 TRY.01 attempt A0: entity trying A1: thing tried Sie müssen vom Bananen-Kuchen kosten! You must of the banana shortcake try! (You must try the banana shortcake.) Example: Annotation Projection Impact of Aliasing Merge Decision Verbs with no English frames Two-Step Approach for Curating Frame Mappings • Step 1: Filter incorrect mappings • Step 2: Group “usage-synonyms’ • Execute approach to create curated frame mappings for three languages Universal Proposition Banks Version 1.0 Public release! https://github.com/System-T/UniversalPropositions/