This paper presents and compares WordNetbased and distributional similarity approaches. The strengths and weaknesses of each approach regarding similarity and relatedness tasks are discussed, and a combination is presented. Each of our methods independently provide the best results in their class on the RG and WordSim353 datasets, and a supervised combination of them yields the best published results on all datasets. Finally, we pioneer cross-lingual similarity, showing that our methods are easily adapted for a cross-lingual task with minor losses.

We present an LSTM approach to deletion-based sentence compression where the task is to translate a sentence into a sequence of zeros and ones, corresponding to token deletion decisions. We demonstrate that even the most basic version of the system, which is given no syntactic information (no PoS or NE tags, or dependencies) or desired compression length, performs surprisingly well: around 30% of the compressions from a large test set could be regenerated. We compare the LSTM system with a competitive baseline which is trained on the same amount of data but is additionally provided with all kinds of linguistic features. In an experiment with human raters the LSTMbased model outperforms the baseline achieving 4.5 in readability and 3.8 in informativeness.

Esta es la versión de autor del artículo publicado en: This is an author produced version of a paper published in: Abstract.Learning style models constitute a valuable tool for improving individual learning by the use of adaptation techniques based on them. In this paper we present how the benefit of considering learning styles with adaptation purposes, as part of the user model, can be extended to the context of collaborative learning as a key feature for group formation. We explore the effects that the combination of students with different learning styles in specific groups may have in the final results of the tasks accomplished by them collaboratively. With this aim, a case study with 166 students of Computer Science has been carried out, from which conclusions are drawn. We also describe how an existing web-based system can take advantage of learning style information in order to form more productive groups. Our ongoing work concerning the automatic extraction of grouping rules starting from data about previous interactions within the system is also outlined. Finally, we present our challenges, related to the continuous improvement of collaboration by the use and dynamic modification of automatic grouping rules.

Users try to articulate their complex information needs during search sessions by reformulating their queries. To make this process more e ective, search engines provide related queries to help users in specifying the information need in their search process. In this paper we propose a customized sequence-to-sequence model for sessionbased query suggestion. In our model, we employ a query-aware a ention mechanism to capture the structure of the session context. is enables us to control the scope of the session from which we infer the suggested next query, which helps not only handle the noisy data but also automatically detect session boundaries. Furthermore we observe that, based on the user query reformulation behavior, within a single session a large portion of query terms is retained from the previously submi ed queries and consists of mostly infrequent or unseen terms that are usually not included in the vocabulary. We therefore empower the decoder of our model to access the source words from the session context during decoding by incorporating a copy mechanism. Moreover, we propose evaluation metrics to assess the quality of the generative models for query suggestion. We conduct an extensive set of experiments and analysis. e results suggest that our model outperforms the baselines both in terms of the generating queries and scoring candidate queries for the task of query suggestion.

Esta es la versión de autor de la comunicación de congreso publicada en: This is an author produced version of a paper published in: Abstract. We describe an approach taken for automatically associating entries from an on-line encyclopedia with concepts in an ontology or a lexical semantic network. It has been tested with the Simple English Wikipedia and WordNet, although it can be used with other resources. The accuracy in disambiguating the sense of the encyclopedia entries reaches 91.11% (83.89% for polysemous words). It will be applied to enriching ontologies with encyclopedic knowledge.