HITS at TAC KBP 2015: Entity Discovery and Linking, and Event Nugget Detection
نویسندگان
چکیده
HITS participated in the English Entity Linking and Discovery (EDL) and Event Nugget Detection (EN task 1) tracks at TAC KBP 2015. Our EDL system introduces a novel, interleaved multitasking approach, which allows interaction between interdependent entity linking subtasks while avoiding the structural and algorithmic complexity of joint models. Out of the eight systems that participated in the monolingual English EDL evaluation, our system is ranked first in linking and second in clustering. HITS also participated in the Event Nugget Detection task. Our approach is based on a structured perceptron which predicts entity mentions, event triggers, and realis jointly. We generalize to unknown triggers with hidden units, a concept used by the frame-semantic parser SEMAFOR.
منابع مشابه
Illinois CCG TAC 2015 Event Nugget, Entity Discovery and Linking, and Slot Filler Validation Systems
This paper describes University of Illinois’s Cognitive Computation Group (UI-CCG)’s submissions for three TAC tracks: Event Nugget Detection/Coreference; Entity Discovery and Linking (EDL); and Slot Filler Validation (SFV). The Event Nugget Detection and Coreference system employs a supervised model for event nugget detection with rich lexical and semantic features while we experiment with bot...
متن کاملEvent Argument Linking and Event Nugget Detection Task: IHMC DISCERN System Report
We present the design and implementation of three variants of an event-detection system called DISCERN, one that used manually created rules (DISCERN-R), one that used machine-learned rules (DISCERN-ML), and one that combined manual and machine-learned rules (DISCERN-C). All three used linguistic resources (VerbNet, CatVar, Semantic Role Labeling, NER, POS tagging, dependency parsing, and coref...
متن کاملCMU System for Entity Discovery and Linking at TAC-KBP 2015
This paper describes CMU’s system for the Tri-lingual Entity Discovery and Linking (TEDL) task at TAC-KBP 2015. Our system is a unified graph-based approach which is able to do concept disambiguation and entity linking simultaneously, leveraging the ontology built on Freebase. The results show that our system achieves competitive results for Chinese and Spanish.
متن کاملUniversity of Washington TAC-KBP 2016 System Description
This document describes the University of Washington’s event extraction system used in the Event Argument Extraction and Linking and Event Nugget Detection tasks of the 2016 TAC KBP competition. This system was composed of three components: Evento, a CRFbased extractor, NomEvent, which makes use a lexicon to build features to identify nominal triggers, and NewsSpike, which uses an unsupervised ...
متن کاملEvent Nugget Detection Task : UMBC systems
This paper described our Event Nugget Detection system that we submitted to the TAC KBP 2016 Event Track. We sent out two runs; UMBC1 and UMBC2. UMBC1 is a sentencelevel classification system based on Convolution Neural Network and applied the probability to select a word as an event nugget. UMBC2 is the classification model trained from our features using Weka and filtered out low confidence p...
متن کامل