Text-Mining Research Group

TAC 2011 multiling pilot overview

The Text Analysis Conference MultiLing Pilotof 2011 posed a multi-lingual summarizationtask to the summarization community, aimingto quantify and measure the performance ofmulti-lingual, multi-document summarizationsystems. The task was to create a 240–250word summary from 10 news texts, describinga given topic. The texts of each topic wereprovided in seven languages (Arabic, Czech,English, French, Greek, Hebrew, Hindi) andeach participant generated summaries for atleast 2 languages. The evaluation of the summarieswas performed using automatic (AutoSummENG,Rouge) and manual processes(Overall Responsiveness score). The participatingsystems were 8, some of which providingsummaries across all languages. Thispaper provides a brief description for the collectionof the data, the evaluation methodology,the problems and challenges faced, andan overview of participation and correspondingresults.

Keywords: Multilingual summarisation evaluation, Text analysis conference

Year: 2012

Download:  Full text 

G. Giannakopoulos

M. El-Haj

B.  Favre

M.  Litvak

Josef Steinberger

E-mail: jstein@kiv.zcu.cz

Josef is an associated professor at the Department of computer science and engineering at the University of West Bohemia in Pilsen, Czech Republic. He is interested in media monitoring and analysis, mainly automatic text summarisation, sentiment analysis and coreference resolution.

V.  Varma

Automatic Text Summarisation
Authors:	Josef Steinberger, Karel Ježek, Michal Campr, Jiří Hynek
Desc.:	Automatic text summarisation using various text mining methods, mainly Latent Semantic Analysis (LSA).