The process of automatically producing a summary from a source text has been intensively investigated in the last decade. This paper presents a new summarization system, developed at Universit¿ de Montr¿al, called sumUM. The system is implemented in SICStus Prolog and Perl, for Sun workstations and Linux machines. It produces short automatic abstracts when long scientific and technical texts are provided as input.
According to the authors, the research was based on the intensive study of manual alignments between sentences of professional abstracts and elements of source documents, and on the exploration of the essential differences between indicative and informative abstracts.
In an evaluation of the sumUM in comparison to other text summarization systems, no differences were observed related to the indicative content, but significant differences were reported related to the informative content. The authors report good performance when compared with other summarization technologies.
I liked the organization of the paper, the subject coverage, and the evaluation report. The bibliography is adequate. Readers interested in computational linguistics will appreciate this practical contribution. The sumUM system is a valuable tool for users interested in text summarization. I recommend this paper to all of them.