Development of Yes/No Arabic Question Answering System

Reading time: 2 minute
...

📝 Original Info

  • Title: Development of Yes/No Arabic Question Answering System
  • ArXiv ID: 1302.5675
  • Date: 2013-02-25
  • Authors: Researchers from original ArXiv paper

📝 Abstract

Developing Question Answering systems has been one of the important research issues because it requires insights from a variety of disciplines,including,Artificial Intelligence,Information Retrieval, Information Extraction,Natural Language Processing, and Psychology.In this paper we realize a formal model for a lightweight semantic based open domain yes/no Arabic question answering system based on paragraph retrieval with variable length. We propose a constrained semantic representation. Using an explicit unification framework based on semantic similarities and query expansion synonyms and antonyms.This frequently improves the precision of the system. Employing the passage retrieval system achieves a better precision by retrieving more paragraphs that contain relevant answers to the question; It significantly reduces the amount of text to be processed by the system.

💡 Deep Analysis

Deep Dive into Development of Yes/No Arabic Question Answering System.

Developing Question Answering systems has been one of the important research issues because it requires insights from a variety of disciplines,including,Artificial Intelligence,Information Retrieval, Information Extraction,Natural Language Processing, and Psychology.In this paper we realize a formal model for a lightweight semantic based open domain yes/no Arabic question answering system based on paragraph retrieval with variable length. We propose a constrained semantic representation. Using an explicit unification framework based on semantic similarities and query expansion synonyms and antonyms.This frequently improves the precision of the system. Employing the passage retrieval system achieves a better precision by retrieving more paragraphs that contain relevant answers to the question; It significantly reduces the amount of text to be processed by the system.

📄 Full Content

Developing Question Answering systems has been one of the important research issues because it requires insights from a variety of disciplines,including,Artificial Intelligence,Information Retrieval, Information Extraction,Natural Language Processing, and Psychology.In this paper we realize a formal model for a lightweight semantic based open domain yes/no Arabic question answering system based on paragraph retrieval with variable length. We propose a constrained semantic representation. Using an explicit unification framework based on semantic similarities and query expansion synonyms and antonyms.This frequently improves the precision of the system. Employing the passage retrieval system achieves a better precision by retrieving more paragraphs that contain relevant answers to the question; It significantly reduces the amount of text to be processed by the system.

Reference

This content is AI-processed based on ArXiv data.

Start searching

Enter keywords to search articles

↑↓
ESC
⌘K Shortcut