Abstract
In text generation, studies on aggregation often focus on the use of connectives to combine short made-up sentences. But connectives restrict the number of units that may be combined at any one time. So, how does information get condensed into fewer units without excessive use of connectives? From a comparison of document and abstract, this reconnaissance study reports on some preferred patterns in aggregation when authors write abstracts for journal articles on biology. The paper also discusses some prerequisites and difficulties anticipated for abstracting systems. More sentences were aggregated without than with the use of an explicit sign, such as a connective or a (semi-)colon.- Anthology ID:
- 2001.jeptalnrecital-long.12
- Volume:
- Actes de la 8ème conférence sur le Traitement Automatique des Langues Naturelles. Articles longs
- Month:
- July
- Year:
- 2001
- Address:
- Tours, France
- Editor:
- Denis Maurel
- Venue:
- JEP/TALN/RECITAL
- SIG:
- Publisher:
- ATALA
- Note:
- Pages:
- 142–151
- Language:
- URL:
- https://aclanthology.org/2001.jeptalnrecital-long.12
- DOI:
- Cite (ACL):
- Choy-Kim Chuah. 2001. Aggregation by Conflation of Quasi-Synonymous Units in Author Abstracting. In Actes de la 8ème conférence sur le Traitement Automatique des Langues Naturelles. Articles longs, pages 142–151, Tours, France. ATALA.
- Cite (Informal):
- Aggregation by Conflation of Quasi-Synonymous Units in Author Abstracting (Chuah, JEP/TALN/RECITAL 2001)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-4/2001.jeptalnrecital-long.12.pdf