NLPComp in TAC 2012 Entity Linking and Slot-Filling

نویسندگان

  • Jian Xu
  • Qin Lu
  • Jie Liu
  • Ruifeng Xu
چکیده

The NLPComp team participated in two TACKBP2012 tasks: Regular Entity Linking and Regular Slot Filling. For the entity linking task, a three-step entity linking system is developed. In the first step, a list of possible candidates are selected. Then the best candidate is identified to decide whether a link exists. In addition, a document clustering algorithm is used to group NIL queries. This system uses the Wikipedia anchor terms to enlarge the number of candidate instances. It then incorporates the topic modelling technique to select features of topical words. However, our system produces a poor answer coverage and the NIL detection system brings significant loss in the final Fscore. For the slot filling task, we developed a system which combines rule-based approach and multiple instance learning technique. In rulebased slot filling, a number of trigger words are collected from the English Wikipedia and the frequency feature is used to select final slot value(s). When extracting slot value(s) using the multiple instance learning technique, 2009 and 2010 KBP slot filling queries and manually annotated answers are used to create named entity pairs which exhibit a particular relation. Then bags of sentences containing named entity pair are extracted from the KBP source. Our system reaches the median level among all the participating systems.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

CUNY BLENDER TAC-KBP2012 Entity Linking System and Slot Filling Validation System

This year the CUNY-BLENDER team participated in the English Entity Linking and Slot Filling Validation tracks. for entity linking, we apply two new techniques, collaborative clustering and query reformulation. For answer validation, we use a logistic regression model trained on within-system and crosssystem features to re-rank the merged answer sets generated by individual systems. In this pape...

متن کامل

Intelius-NYU TAC-KBP2012 Cold Start System

This paper describes the Intelius-NYU 2012 system for the KBP Cold Start task. The Cold Start task can be decomposed into two subtasks: slot filling and entity linking. For slot filling, we focus on the adaptation of the NYU 2011 regular slot filling system to the Cold Start task. For entity linking, we apply Intelius‘s commercial conflation engine to link person and organization entities with ...

متن کامل

IIIT Hyderabad at TAC 2012

In this paper, we report our participation in Knowledge Base Population at TAC 2012. We adopted an Information Retrieval based approach for the Entity Linking and Slot Filling tasks. In Entity Linking we identify potential nodes from the Knowledge Base and then identify the mapping node using tf-idf similarity. We achieved very good performance in the Entity Linking task. For Slot Filling task ...

متن کامل

The TALP participation at TAC-KBP 2013

This document describes the work performed by the Universitat Politècnica de Catalunya (UPC) in its second participation at TAC-KBP 2013 in both the Entity Linking and the Slot Filling tasks.

متن کامل

HLTCOE Approaches to Knowledge Base Population at TAC 2009

The HLTCOE participated in the entity linking and slot filling tasks at TAC 2009. A machine learning-based approach to entity linking, operating over a wide range of feature types, yielded good performance on the entity linking task. Slot-filling based on sentence selection, application of weak patterns and exploitation of redundancy was ineffective in the slot filling task.

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2012