Abstract
Information and Communication Technologies (ICT) have revolutionised the tourism domain, providing a wide set of new services for tourists and tourism businesses. Both tourists and tourism businesses use dedicated tourism platforms to search and share information generating, constantly, new tourism crowdsourced data. This crowdsourced information has a huge influence in tourist decisions. In this context, the paper proposes a stream recommendation engine supported by crowdsourced information, adopting Stochastic Gradient Descent (SGD) matrix factorisation algorithm for rating prediction. Additionally, we explore different (i) profiling approaches (hotel-based and theme-based) using hotel multi-criteria ratings, location, value for money (VfM) and sentiment value (StV); and (ii) post-recommendation filters based on hotel location, VfM and StV. The main contribution focusses on the application of post-recommendation filters to the prediction of hotel guest ratings with both hotel and theme multi-criteria rating profiles, using crowdsourced data streams. The results show considerable accuracy and classification improvement with both hotel-based and theme-based multi-criteria profiling together with location and StV post-recommendation filtering. While the most promising results occur with the hotel-based version, the best theme-based version shows a remarkable memory conciseness when compared with its hotel-based counterpart. This makes this theme-based approach particularly appropriate for data streams. The abstract completely needs to be rewritten. It does not provide a clear view of the problem and its solutions the researchers proposed. In addition, it should cover five main elements, introduction, problem statement, methodology, contributions and results. Done.

Abstract
Wiki-based crowdsourced repositories have increasingly become an important source of information for users in multiple domains. However, as the amount of wiki-based data increases, so does the information overloading for users. Wikis, and in general crowdsourcing platforms, raise trustability questions since they do not generally store user background data, making the recommendation of pages particularly hard to rely on. In this context, this work explores scalable multi-criteria profiling using side information to model the publishers and pages of wiki-based crowdsourced platforms. Based on streams of publisher-page-review triads, we have modelled publishers and pages in terms of quality and popularity using different criteria and user-page-view events collected via a wiki platform. Our modelling approach classifies statistically, both page-review (quality) and pageview (popularity) events, attributing an appropriate rating. The quality-related information is then merged employing Multiple Linear Regression as well as a weighted average. Based on the quality and popularity, the resulting page profiles are then used to address the problem of recommending the most interesting wiki pages per destination to viewers. This paper also explores the parallelisation of profiling and recommendation algorithms using wiki-based crowdsourced distributed data repositories as data streams via incremental updating. The proposed method has been successfully evaluated using Wikivoyage, a tourism crowdsourced wiki-based repository.

Abstract
Nowadays, with the widely usage of on-line stream video platforms, the number of media resources available and the volume of crowd-sourced feedback volunteered by viewers is increasing exponentially. In this scenario, the adoption of recommendation systems allows platforms to match viewers with resources. However, due to the sheer size of the data and the pace of the arriving data, there is the need to adopt stream mining algorithms to build and maintain models of the viewer preferences as well as to make timely personalised recommendations. In this paper, we propose the adoption of optimal individual hyper-parameters to build more accurate dynamic viewer models. First, we use a grid search algorithm to identify the optimal individual hyper-parameters (IHP) and, then, use these hyper-parameters to update incrementally the user model. This technique is based on an incremental learning algorithm designed for stream data. The results show that our approach outperforms previous approaches, reducing substantially the prediction errors and, thus, increasing the accuracy of the recommendations. © 2019 for this paper by its authors.

Abstract
Crowdsourced repositories have become an increasingly important source of information for users and businesses in multiple domains. Everyday examples of tourism crowdsourcing platforms focusing on accommodation, food or travelling in general, influence consumer behaviour in modern societies. These repositories, due to their intrinsic openness, can strongly benefit from independent data quality modelling mechanisms. In this context, building trust & reputation models of contributors and storing crowdsourced data using distributed ledger technology allows not only to ascertain the quality of crowdsourced contributions, but also ensures the integrity of the built models. This paper presents a survey on distributed trust & reputation modelling using blockchain technology and, for the specific case of tourism crowdsourcing platforms, discusses the open research problems and identifies future lines of research. 2019 The Authors. Published by Elsevier B.V.

Abstract
We studied the problem of how to identify the most impactful papers of a scientific field, for longitudinal bibliometric analyses or systematic literature reviews' purposes. We show that using raw citation counts, the most popular approach, it is not suitable to compare papers from different periods. Other approaches, such as the use of normalized citations by the paper's exposure time, or by the annual average citations of the area, although improving the selection quality, do not lead to sufficiently homogeneous results in terms of citation counts and number of papers published per year. As an alternative, we propose a percentile citation-based method and compare it to the commonly used approaches, for the Top100, and the Top500 in a sample of 25144 papers. This sample was collected from the Scopus database, by selecting the top 10% sources titles in the period 1987-2015 in the Archaeology field. Our results show that the choice of the right normalization metric to be used in the ranking of the impact of the papers is crucial, since it may privilege certain periods, while neglecting others. Based on our results, we argue that this does not happen with our approach, the percentile method.

Abstract
Citation analysis has been used to compare researchers, fields, institutions and countries. However, not much has been done to compare citations of papers belonging to different databases and published in different years. This comparison could play a relevant role in many systematic literature reviews concerned with the growth, development, and changes of a particular scientific subject. This study aims to examine whether we can use the percentile approach to compare the number of citations from papers in different databases. We argue that this method can convert citations from different databases when there are same articles belonging to more than one database. We apply the method on Thomson Reuters' Web of Science and Elsevier's Scopus databases because they are the leading databases of scholarly impact. In this study we use two different Scopus subject area: Engineering - Industrial and Manufacturing Engineering; and Arts and Humanities -Archaeology. The analysis comprises articles published for the time period 1987-2017, of journals in the Scopus top 10%, corresponding to approximately 152,000 papers.