Abstract
The internet is acknowledge as the main tourism communication medium and business facilitator. However, its functionality in this sector has been limited to e-commerce and focused on meeting the demand, thus underusing its potential as an essential tool for offer development, through the opportunities created by e-business. Since tourism is an eminently relational activity that strengthens itself from the sum of the joint efforts of its components, but oten fragmented and dispersed, this article advocates the adoption of online interorganizational collaboration platforms, which provides na environment for interactions, cooperation, and knowledge sharing amongst the social actors of tourist destinations. The proposal is based on the methodology of discourse analysis of extant literature on the internet economy and social network theory in tourism, exemplifying the advantages and difficulties that may arise from such a strategy. Recognizing that the available literature on this subject is scarce, three questions are also identified that can be tackled by future research.

Abstract
Multi-label classification is a methodology that tries to solve classification problems where multiple classes are associated with each data example. Data streams pose new challenges to this methodology caused by the massive amounts of structured data production. In fact, most of the existent batch mode methods may not support this condition. Therefore, this paper proposes two multi-label classification methods based on rule and ensembles learning from continuous flow of data. These methods are derived from a multi-target regression algorithm. The main contribution of this work is the rule specialization for subsets of class labels, instead of the usual local (individual models for each output) or a global (one model for all outputs) methods. Prequential evaluation was conducted where global, local and subset operation modes were compared against other online classifiers found in the literature. Six real-world data sets were used. The evaluation demonstrated that the subset specialization presents competitive performance, when compared to local and global approaches and online classifiers found in the literature.

Abstract
This paper describes the development of a Co-training (semi-supervised approach) method that uses multiple learners for single target regression on data streams. The experimental evaluation was focused on the comparison between a realistic supervised scenario (all unlabelled examples are discarded) and scenarios where unlabelled examples are used to improve the regression model. Results present fair evidences of error measure reduction by using the proposed Co-training method. However, the error reduction still is relatively small.

Abstract
Well logs are records of petro-physical data acquired along a borehole, providing direct information about what is in the subsurface. The data collected by logging wells can have significant economic consequences in oil and gas exploration, not only because it has a direct impact on the following decisions, but also due to the subsequent costs inherent to drilling wells, and the potential return of oil deposits. These logs frequently present gaps of varied sizes in the sensor recordings, that happen for diverse reasons. These gaps result in less information used by the interpreter to build the stratigraphic models, and consequently larger uncertainty regarding what will be encountered when the next well is drilled. The main goal of this work is to compare Gradient Tree Boosting, Random Forests, Artificial Neural Networks, and three algorithms of Linear Regression on the prediction of the gaps in well log data. Given the logs from a specific well, we use the intervals with complete information as the training data to learn a regression model of one of the sensors for that well. The algorithms are compared with each other using a few individual example wells with complete information, on which we build artificial gaps to cross validate the results. We show that the ensemble algorithms tend to perform significantly better, and that the results hold when addressing the different examples individually. Moreover, we performed a grid search over the ensembles parameters space, but did not find a statistically significant difference in any situation.

Abstract
In this paper, we investigate two variants of association rules for preference data, Label Ranking Association Rules and Pairwise Association Rules. Label Ranking Association Rules (LRAR) are the equivalent of Class Association Rules (CAR) for the Label Ranking task. In CAR, the consequent is a single class, to which the example is expected to belong to. In LRAR, the consequent is a ranking of the labels. The generation of LRAR requires special support and confidence measures to assess the similarity of rankings. In this work, we carry out a sensitivity analysis of these similarity-based measures. We want to understand which datasets benefit more from such measures and which parameters have more influence in the accuracy of the model. Furthermore, we propose an alternative type of rules, the Pairwise Association Rules (PAR), which are defined as association rules with a set of pairwise preferences in the consequent. While PAR can be used both as descriptive and predictive models, they are essentially descriptive models. Experimental results show the potential of both approaches.

Abstract
Neural networks have been applied as a machine learning tool in many different areas. Recently, they have gained increased attention with what is now called deep learning. Neural networks algorithms have several parameters that need to be tuned in order to maximize performance. The definition of these parameters can be a difficult, extensive and time consuming task, even for expert users. One approach that has been successfully used for algorithm and parameter selection is metalearning. Metalearning consists in using machine learning algorithm on (meta)data from machine learning experiments to map the characteristics of the data with the performance of the algorithms. In this paper we study how a metalearning approach can be used to obtain a good set of parameters to learn a neural network for a given new dataset. Our results indicate that with metalearning we can successfully learn classifiers from past learning tasks that are able to define appropriate parameters.