A summarizer for different tasks involving aspects related with the language, the media and the domain of the document to be summarized.

Currently on-line summarization of generic or scientific textual or spontaneous speech documents in English.

The summarizer is developed in Perl, and a specific wrapper is developed when some component using different technology is included (Freeling, TNT, YAMCHA, …).

The summarizer requires a linguistic preprocess. For Catalan and Spanish Freeling and euroWordnet are used, while TNT, WordNet and YAMCHA are used to process English spontaneous speech documents.

Linguistic preprocessing; Lexical chain based summarizer (includes: Discourse Marker annotator, Text Tiler, Lexical Chainer and chunkSum).

The FEMsum was developed for Maria Fuentes’ Phd thesis within the framework of the HERMES, ALIADO and CHIL projects.

• Maria Fuentes. “A Flexible Multitask Summarizer for Documents from Different Media, Domain, and Language”. Ph.D. Thesis on Artificial Intelligence, Advisor Horacio Rodríguez. March, 2008.
• Maria Fuentes, Edgar Gonzàlez, Horacio Rodríguez, Jordi Turmo, Laura Alonso. “Summarizing Spontaneous Speech Using General Text Properties”. In Proceedings of the Crossing Barriers in Text Summarization Research Workshop held in conjunction with RANLP, Borovets, Bulgary, September 2005.
• Maria Fuentes, Edgar Gonzàlez, and Horacio Rodríguez. “Resumidor de noticies en català del projecte Hermes”. II Congrés d’Enginyeria en Llengua Catalana, Andorra, 2004.
• Laura Alonso, Maria Fuentes. “Integrating Cohesion and Coherence for Text Summarization”. In Proceedings of the EACL’03 Student Session, Budapest, Hungary, April 2003.
• Maria Fuentes, Horacio Rodríguez. “Using cohesive properties of text for Automatic Summarization”. In Proceedings of the Primeras Jornadas de Tratamiento y Recuperación de Información, Valencia, Spain, 2002.