Center for Language Engineering
ناشر
Pakistan
ملک
Lahore
شہر
09-11-2012
تاریخِ آغاز
10-11-2012
تاریخِ اختتام


تلخیص
The conference on the climate change, UNFCCC 2010, took place from 29 November to 10 December 2010 in Cancun, Mexico. This paper presents an analysis of the opening speeches by various countries at the conference, combined with the statistics of the countries regarding their socioeconomic indicators and memberships of different climate treaties. A central objective is to compare different sources of the information that reflect the underlying complex system where there are obvious and less obvious relationships between the rhetoric and some aspects of reality. At the level of argumentation, we are interested in the occurrence of topics related to the climate change, i.e., whether some topics are mentioned or avoided in the speeches. The recognition of the topics is based on a semi-automatic term selection process that provides the input for the subsequent steps of the analysis. The data preparation process includes optical character recognition, machine translation and approximate string matching. We assume that the collection of terms serves as a relevant set of features that reflect the content of the speeches. These text-based features are then compared with the country statistics. The basic hypothesis is that there is a detectable but complex relationship between the content of the speeches and known facts. The most important contributions in this paper are the formulation of the basic questions and the overall hypothesis, an analysis of the relationships between the countries as well as between the topics and indicators, and the qualitative analysis of the results.

Yasir Mehmood, Timo Honkela. (2012) Comparing Talks, Realities and Concerns over the Climate Change: Comparing Texts with Numerical and Categorical Data, Conference on Language and Technology 2012.
  • Viewed 1493
  • Downloads 124