Machine learning approaches for intelligent text processing are actively studied by researchers in natural language processing, information retrieval and related fields. While supervised methods usually attain much better performance than unsupervised methods, they also require annotated data which is not always available or easy to obtain. Hence, we encourage the submission of supervised, unsupervised or hybrid methods for intelligent text processing tasks. Methods studying alternative learning paradigms, e.g. semi-supervised learning, weakly-supervised learning, zero-shot learning but also transfer learning, are very welcome as well.
This sessions covers three research areas: natural language processing, computational linguistics and information retrieval. We invite the submission of papers on original and unpublished research in all aspects of automated and intelligent text processing, including but not limited to the following topics:
Information retrieval, information extraction
Question answering
Machine learning in NLP
Machine translation
Morphology, syntax and semantics
Resources and evaluation
Representation of linguistic knowledge
Sentiment analysis and opinion mining
Social media
Summarization
Text Mining
Text categorization and clustering
Word sense disambiguation
NEW! Papers
Andrei Butnaru, Radu Tudor IonescuFrom Image to Text Classification: A Novel Approach based on Clustering Word Embeddings
Paper: • Slides:
Mihai Masala, Stefan Ruseti, Traian RebedeaSentence selection with neural networks over string kernels
Paper: • Slides:
Mouna Kamel, Nathalie Aussenac-Gilles, Cécile Fabre, Adel Ghamnia, Cassia TrojahnA Distant Learning Approach for Extracting Hypernym Relations from Wikipedia Disambiguation Pages
Paper: • Slides:
Marius Popescu, Cristian Grozea, Radu Tudor IonescuHASKER: An efficient algorithm for string kernels. Application to polarity classification in various languages
Paper: • Slides:
Ismail Badache, Sébastien Fournier, Adrian-Gabriel ChifuHarnessing Ratings and Aspect-Sentiment to Estimate Contradiction Intensity in Temporal-Related Reviews Paper: • Slides:
Marc Franco-Salvador, Greg Kondrak, Paolo RossoBridging the Native Language and Language Variety Identification Tasks Paper: • Slides:
Josiane Mothe, Mahdi WashhaPredicting the Best System Parameter Configuration: the (Per Parameter Learning) PPL method
Paper: • Slides:
Laura Franzoi, Andrea SgarroLinguistic classification: T-norms, fuzzy distances and fuzzy distinguishabilities
Paper: • Slides:
Important Dates
Submission of Papers: 20th of April, 2017 (extended)*
Notification of Acceptance: 10th of May, 2017 (extended)
Camera Ready Version: 26th of May, 2017
Author Registration Deadline: 2nd of June, 2017
Conference Dates: 6-8 September, 2017
*There will be no further extension provided after this date.
Open Access Publication
Accepted papers will be published in the KES 2017 conference proceedings. The KES conference is ranked B by the CORE Australian Association. The conference proceedings will appear in Elsevier’s Procedia Computer Science. The proceedings will appear online, on ScienceDirect, within 8 weeks of acceptance of the final manuscripts via the conference organizer and are freely available to all users.
Important Note:
Every paper must have at least one author who has registered for the conference with payment by the Early Registration Deadline for the paper to appear in the proceedings. Any paper that does not have at least one author who has registered and paid by the Early / Author Registration Deadline will be withdrawn from the proceedings.
Paper Format
Submissions for the invited session must be made as complete papers (there is no abstract submission stage) submitted as PDF documents through the PROSE online submission and review system. Full papers should be detailed academic articles in conventional format. The guide length for full papers is 8 to 10 pages (maximum).
Guidance notes for the preparation of Full Papers is available
An MS Word template is available (5.3Mb .docm file)
A LaTeX template is available (7Mb .zip archive)
The paper format as a PDF document is available
Please consult important FAQs about document preparation to be found
For a paper to be published in the Procedia proceedings:
the maximum length of 10 pages must not be exceeded
the paper must be presented at the conference
It is the author's responsibility to ensure that their paper does not contain any errors. Also, kindly note that Elsevier will publish what they receive so it is important that the authors submit the final version of their papers. Proofs will not be sent to authors at any time during production.
Marseille is now France's largest city on the Mediterranean coast and the largest port for commerce, freight and cruise ships. The city was European Capital of Culture in 2013. Marseille is listed as a major centre of art and history. The city has many museums and galleries and there are many ancient buildings and churches of historical interest.
Contact Details
For details please send an e-mail to:
adrian.chifu@lsis.org
Adrian-Gabriel CHIFU
Aix-Marseille Université (FEG), LSIS - France