Neto, Joel Larocca and Santos, Alexandre D. and Kaestner, Celso A.A. and Freitas, Alex A. and Nievola, Julio C.
A Trainable Algorithm for Summarizing News Stories.
In: Proc. PKDD'2000 Workshop on Machine Learning and Textual Information Access.
(Full text available)
This work proposes a trainable system for summarizing news and obtaining an approximate
argumentative structure of the source text. To achieve these goals we use several techniques
and heuristics, such as detecting the main concepts in the text, connectivity between sentences,
occurrence of proper nouns, anaphors, discourse markers and a binary-tree representation (due
to the use of an agglomerative clustering algorithm). The proposed system was evaluated on a
set of 800 documents.
- Depositors only (login required):
Downloads per month over past year