ag-sc / QueGG-web

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Run QA system on your machine

This page provides instructions on how to run the QA system on your machine. The home page of the hackathon can be found here

Install QA system

Install [docker] (https://docs.docker.com/engine/install/)

  1. Download the image
docker pull agsc/quegg-web:latest
  1. Run the image as a container.
docker run -p "8089:8089" -e "QUEGG_ALLOW_UPLOADS=true" agsc/quegg-web:latest

Go to http://localhost:8089/quegg/ and the interface will be shown on your browser. It will initially be empty, a minimal example to get data into the running instance would be:

Add questions to the QA system

  1. Download the file containing lexical entries. The file nounppframe.csv shows an example of the lexical entry.
wget -O nounppframe.csv https://raw.githubusercontent.com/ag-sc/QueGG-web/main/example/nounppframe.csv

Post the file

curl -X "POST" -F "file=@nounppframe.csv" "http://localhost:8089/quegg/import"      

Add more questions to the QA system

a) add lexical entry at Google XSL sheet. See the guideline of writing a lexical entry for a grammar type.
b) download the Google XSL sheet as csv. File>Download>Comma-separated values [.csv, current sheet].
c) repeat step 3. The questions of newly added lexical entry will be visible in QA system.

Please use the following citation:

@inproceedings{Buono-LREC2020,
	title = {{Generating Grammars from lemon lexica for Questions Answering over Linked Data: a Preliminary Analysis}},
	author = {Viktoria Benz, Philipp Cimiano, Mohammad Fazleh Elahi, Basil Ell},
	booktitle = {In: NLIWOD workshop at ISWC 2020},
	pages = {40–55},
	year = {2020},
	link = {http://ceur-ws.org/Vol-2722/nliwod2020-paper-2.pdf}
}

Developers

  • Frank Grimm
  • Mohammad Fazleh Elahi

Supervisors:

  • Dr. Philipp Cimiano
  • Dr. Basil Ell

About


Languages

Language:Java 72.6%Language:JavaScript 17.4%Language:HTML 5.1%Language:CSS 2.0%Language:Makefile 1.0%Language:Shell 1.0%Language:Dockerfile 0.8%