Guest editors: Dan Roth – University of Illinois, Urbana Champaign, Roberto Basili – University of Roma, Tor Vergata.
Call Deadline: July 20, 2017
The second 2017 issue of IJCol will be a special issue devoted to “Language and Learning Machines”. Learning is a fundamental ability with respect to the acquisition and use of natural languages within diverse and geographically distributed communities. The ability to apply learning algorithms has been the major cause behind the renaissance of a variety of empirical methods in computational linguistics since the late ‘90s and it is now amplified by the huge increase in the computational power of modern computing, and the availability of data and cheap storage. The re-emergence of neural architectures and their use in this context provides a key instance of this trend. However, the nature, structure and variety of the processes required to support natural language understanding and the generality of their application is still debated, as little is known, for example, about the nature and validity of these general strategies in language learning. Open questions abound and range from representational issues to those that bridge the gap between advancing natural language engineering and better understanding meaningful linguistic abstractions.
Papers are welcome reporting on all these open issues through the lens of novel on-going research as well as survey papers, position papers or reviews and project reports. Papers devoted to phenomena and specialties related to the Italian language will be of a particular but not exclusive interest.
The topics include but are not limited to:
- Formal Machine Learning methods for Language Processing
- Lexical Learning: distributional methods, word embeddings and neural language models
- Probabilistic and neural models of grammatical and semantic phenomena
- On-line learning, incremental learning, reinforcement learning for natural languages
- Learning with minimal and indirect supervision
- Active learning strategies and autonomous learning in talking agents and machines
- Advanced learning over visual and linguistic input
- Semantic Inference and machine learning
- Pragmatic approaches to learning for NLP
- Cognitive approaches to language learning
- Representation Learning for NLP
- Neural architectures for NLP tasks: from MLPs to recursive NNs
- Scalability of ML over big unstructured data streams
- Linguistically motivated Information Access based on ML
- Inductive methods over Social Web data: computational models of sentiment and emotions
- Learning methods for Human Profiling in Social Web
We welcome manuscripts written in English and Italian. Instructions for manuscript formatting and submission are at the IJCoL page: http://www.ai-lc.it/it/rivista/processo-editoriale/