Quality control, validation and user feedback of the European Flood Alert System (EFAS)

The quality control, validation and verification of the European Flood Alert System (EFAS) are described. EFAS is designed as a flood early warning system at pan-European scale, to complement national systems and provide flood warnings more than 2 days before a flood. On average 20–30 alerts per year are sent out to the EFAS partner network which consists of 24 National hydrological authorities responsible for transnational river basins. Quality control of the system includes the evaluation of the hits, misses and false alarms, showing that EFAS has more than 50% of the time hits. Furthermore, the skills of both the meteorological as well as the hydrological forecasts are evaluated, and are included here for a 10-year period. Next, end-user needs and feedback are systematically analysed. Suggested improvements, such as real-time river discharge updating, are currently implemented.

An investigation into the viability of LibraryThing for promotional and user engagement purposes in libraries

Structured abstract: Purpose: LibraryThing is a Web 2.0 tool allowing users to catalogue books using data drawn from sources such as Amazon and the Library of Congress and has facilities such as tagging and interest groups. This study evaluates whether LibraryThing is a valuable tool for libraries to use for promotional and user engagement purposes. Methodology: This study used a sequential mixed methods 3 phase design: (1) the identification of LibraryThing features for user engagement or promotional purposes, (2) exploratory semi-structured interviews (3) a questionnaire. Findings: Several uses of LibraryThing for promotional and user engagement purposes were identified. The most popular reason libraries used LibraryThing was to promote the library or library stock, with most respondents using it specifically to highlight collections of books. Monitoring of patron usage was low and many respondents had not received any feedback. LibraryThing was commonly reported as being easy to use, remotely accessible, and having low cost, whilst its main drawbacks were the 200 book limit for free accounts, and it being a third-party site. The majority of respondents felt LibraryThing was a useful tool for libraries. Practical implications: LibraryThing has most value as a promotional tool for libraries. Libraries should actively monitor patron usage of their LibraryThing account or request user feedback to ensure that LibraryThing provides a truly valuable service for their library. Orginality : There is little research on the value of LibraryThing for libraries, or librarians perceptions of LibraryThing as a Web 2.0 tool.

Semi-supervised learning for relevance feedback on image retrieval tasks

Relevance feedback approaches have been established as an important tool for interactive search, enabling users to express their needs. However, in view of the growth of multimedia collections available, the user efforts required by these methods tend to increase as well, demanding approaches for reducing the need of user interactions. In this context, this paper proposes a semi-supervised learning algorithm for relevance feedback to be used in image retrieval tasks. The proposed semi-supervised algorithm aims at using both supervised and unsupervised approaches simultaneously. While a supervised step is performed using the information collected from the user feedback, an unsupervised step exploits the intrinsic dataset structure, which is represented in terms of ranked lists of images. Several experiments were conducted for different image retrieval tasks involving shape, color, and texture descriptors and different datasets. The proposed approach was also evaluated on multimodal retrieval tasks, considering visual and textual descriptors. Experimental results demonstrate the effectiveness of the proposed approach.

SaveRF: Towards efficient relevance feedback search

In multimedia retrieval, a query is typically interactively refined towards the ‘optimal’ answers by exploiting user feedback. However, in existing work, in each iteration, the refined query is re-evaluated. This is not only inefficient but fails to exploit the answers that may be common between iterations. In this paper, we introduce a new approach called SaveRF (Save random accesses in Relevance Feedback) for iterative relevance feedback search. SaveRF predicts the potential candidates for the next iteration and maintains this small set for efficient sequential scan. By doing so, repeated candidate accesses can be saved, hence reducing the number of random accesses. In addition, efficient scan on the overlap before the search starts also tightens the search space with smaller pruning radius. We implemented SaveRF and our experimental study on real life data sets show that it can reduce the I/O cost significantly.

Automated illustration of multimedia stories

Submitted in part fulfillment of the requirements for the degree of Master in Computer Science

Mobile image processing to support educational geometry games

The mobile IT era is here, it is still growing and expanding at a steady rate and, most of all, it is entertaining. Mobile devices are used for entertainment, whether social through the so-called social networks, or private through web browsing, video watching or gaming. Youngsters make heavy use of these devices, and even small children show impressive adaptability and skill. However not much attention is directed towards education, especially in the case of young children. Too much time is usually spent in games which only purpose is to keep children entertained, time that could be put to better use such as developing elementary geometric notions. Taking advantage of this pocket computer scenario, it is proposed an application geared towards small children in the 6 – 9 age group that allows them to consolidate knowledge regarding geometric shapes, forming a stepping stone that leads to some fundamental mathematical knowledge to be exercised later on. To achieve this goal, the application will detect simple geometric shapes like squares, circles and triangles using the device’s camera. The novelty of this application will be a core real-time detection system designed and developed from the ground up for mobile devices, taking into account their characteristic limitations such as reduced processing power, memory and battery. User feedback was be gathered, aggregated and studied to assess the educational factor of the application.

Social-media monitoring for cold-start recommendations

Generating personalized movie recommendations to users is a problem that most commonly relies on user-movie ratings. These ratings are generally used either to understand the user preferences or to recommend movies that users with similar rating patterns have rated highly. However, movie recommenders are often subject to the Cold-Start problem: new movies have not been rated by anyone, so, they will not be recommended to anyone; likewise, the preferences of new users who have not rated any movie cannot be learned. In parallel, Social-Media platforms, such as Twitter, collect great amounts of user feedback on movies, as these are very popular nowadays. This thesis proposes to explore feedback shared on Twitter to predict the popularity of new movies and show how it can be used to tackle the Cold-Start problem. It also proposes, at a finer grain, to explore the reputation of directors and actors on IMDb to tackle the Cold-Start problem. To assess these aspects, a Reputation-enhanced Recommendation Algorithm is implemented and evaluated on a crawled IMDb dataset with previous user ratings of old movies,together with Twitter data crawled from January 2014 to March 2014, to recommend 60 movies affected by the Cold-Start problem. Twitter revealed to be a strong reputation predictor, and the Reputation-enhanced Recommendation Algorithm improved over several baseline methods. Additionally, the algorithm also proved to be useful when recommending movies in an extreme Cold-Start scenario, where both new movies and users are affected by the Cold-Start problem.

Machine Learning and Clinical Text. Supporting Health Information Flow

Fluent health information flow is critical for clinical decision-making. However, a considerable part of this information is free-form text and inabilities to utilize it create risks to patient safety and cost-­effective hospital administration. Methods for automated processing of clinical text are emerging. The aim in this doctoral dissertation is to study machine learning and clinical text in order to support health information flow.First, by analyzing the content of authentic patient records, the aim is to specify clinical needs in order to guide the development of machine learning applications.The contributions are a model of the ideal information flow,a model of the problems and challenges in reality, and a road map for the technology development. Second, by developing applications for practical cases,the aim is to concretize ways to support health information flow. Altogether five machine learning applications for three practical cases are described: The first two applications are binary classification and regression related to the practical case of topic labeling and relevance ranking.The third and fourth application are supervised and unsupervised multi-class classification for the practical case of topic segmentation and labeling.These four applications are tested with Finnish intensive care patient records.The fifth application is multi-label classification for the practical task of diagnosis coding. It is tested with English radiology reports.The performance of all these applications is promising. Third, the aim is to study how the quality of machine learning applications can be reliably evaluated.The associations between performance evaluation measures and methods are addressed,and a new hold-out method is introduced.This method contributes not only to processing time but also to the evaluation diversity and quality. The main conclusion is that developing machine learning applications for text requires interdisciplinary, international collaboration. Practical cases are very different, and hence the development must begin from genuine user needs and domain expertise. The technological expertise must cover linguistics,machine learning, and information systems. Finally, the methods must be evaluated both statistically and through authentic user-feedback.

IPTV-palvelun pilotointi

Diplomityössä kehitettiin teleoperaattorille IPTV-palvelu (Internet Protocol Television) tietyillä rajatuilla ominaisuuksilla, kehitettiin IPTV-palvelulle uusi käyttöliittymä ja pilotoitiin uutta palvelua. Pilotoinnin tarkoituksena oli saada kokemusta uudentyyppisten tv-palvelujen käyttöönotosta. Pilotoinnin lopuksi koottiin pilottikäyttäjiltä saatu palaute ja analysoitiin palvelua sen perusteella. Työn taustatieto-osuudessa määritellään IPTV ja siihen tyypillisimmin kuuluvat palvelut. Lisäksi selvitetään, millaisia ovat IPTV:ssä käytetyt tekniikat ja verkkoarkkitehtuuri. Taustatiedoissa esitellään myös palvelun käyttäjän kokeman laadun parametrit ja kerrosmalli. Samalla tuodaan esiin menetelmät erityisesti kuvanlaadun ja käytettävyyden parantamiseksi. Opittavuus ja helppokäyttöisyys sekä viihteellisyys ja miellyttävyys on tunnistettu tärkeimmiksi käytettävyyden osa-alueiksi elinkaarensa alkupäässä olevissa IPTV-palveluissa. Diplomityön toteutusosuudessa esitetään työn lähtökohtana ollut vaatimusmäärittely ja sen vaikutus työn rajaukseen. Tämän jälkeen esitellään kehitetyn IPTV-palvelun ominaisuudet, joista tärkeimpinä tuodaan esiin tallennus- ja ajansiirtotoiminnot. Toteutusosuudessa selvitetään myös teleoperaattorin IPTV-verkon rakenne ja toiminta. Seuraavaksi käydään läpi alkuperäisestä käyttöliittymästä tehty analyysi ja analyysin perusteella toteutettu uusi käyttöliittymä, jossa merkittävimmät parannukset on tehty navigaatiorakenteeseen. Lopuksi selvitetään pilotoinnista kerätyt tulokset. Tuloksien perusteella voidaan nähdä, että ajansiirtotoiminnot riittävät takaamaan palvelun omaksumisen, mutta uudet toiminnot vaativat riittävän opastuksen. Tulokset osoittavat myös, että palvelun käytettävyys uudella käyttöliittymällä on tavoitetasolla, vaikkakin toimintojen viiveet heikentävät käyttökokemuksen miellyttävyyttä.

SharePoint-sovellusten käytettävyys ja sen kehittäminen

Tässä työssä tutkitaan SharePoint-sovellusten käytettävyyden kehittämistä. Käytettävyyden kehittämiselle etsitään keinoja kirjallisuudessa esitetyistä teoreettisista keinoista. Lisäksi selvitetään monipuolisten Internet-sovellusten käytön vaikutuksia SharePoint-sovellusten käytettävyyteen. Tässä työssä toteutetaan SharePoint-ohjelmisto, joka sisältää kaksi erilaista käyttöliittymää. Näitä käyttöliittymiä vertaillaan keskenään käytettävyyden näkökulmasta ja havaitaan, että käyttäjäpalautteen ja monipuolisten Internet-sovellusten ansiosta saadaan toteutettua laadukkaampaa käytettävyyttä. Työssä tehdään kyselytutkimus SharePointia käyttäville yrityksille, jossa selvitetään SharePoint-sovellusten käytettävyyden merkitystä ja laatua. Kyselyyn saadaan vain vähän vastauksia, jonka takia sen pohjalta ei pystytä tekemään yleistäviä johtopäätöksiä.

Method for automatic detection of wheezing in lung sounds

The present report describes the development of a technique for automatic wheezing recognition in digitally recorded lung sounds. This method is based on the extraction and processing of spectral information from the respiratory cycle and the use of these data for user feedback and automatic recognition. The respiratory cycle is first pre-processed, in order to normalize its spectral information, and its spectrogram is then computed. After this procedure, the spectrogram image is processed by a two-dimensional convolution filter and a half-threshold in order to increase the contrast and isolate its highest amplitude components, respectively. Thus, in order to generate more compressed data to automatic recognition, the spectral projection from the processed spectrogram is computed and stored as an array. The higher magnitude values of the array and its respective spectral values are then located and used as inputs to a multi-layer perceptron artificial neural network, which results an automatic indication about the presence of wheezes. For validation of the methodology, lung sounds recorded from three different repositories were used. The results show that the proposed technique achieves 84.82% accuracy in the detection of wheezing for an isolated respiratory cycle and 92.86% accuracy for the detection of wheezes when detection is carried out using groups of respiratory cycles obtained from the same person. Also, the system presents the original recorded sound and the post-processed spectrogram image for the user to draw his own conclusions from the data.

Item-level life-cycle model for maintenance network – from cost to additional value

The importance of industrial maintenance has been emphasized during the last decades; it is no longer a mere cost item, but one of the mainstays of business. Market conditions have worsened lately, investments in production assets have decreased, and at the same time competition has changed from taking place between companies to competition between networks. Companies have focused on their core functions and outsourced support services, like maintenance, above all to decrease costs. This new phenomenon has led to increasing formation of business networks. As a result, a growing need for new kinds of tools for managing these networks effectively has arisen. Maintenance costs are usually a notable part of the life-cycle costs of an item, and it is important to be able to plan the future maintenance operations for the strategic period of the company or for the whole life-cycle period of the item. This thesis introduces an itemlevel life-cycle model (LCM) for industrial maintenance networks. The term item is used as a common definition for a part, a component, a piece of equipment etc. The constructed LCM is a working tool for a maintenance network (consisting of customer companies that buy maintenance services and various supplier companies). Each network member is able to input their own cost and profit data related to the maintenance services of one item. As a result, the model calculates the net present values of maintenance costs and profits and presents them from the points of view of all the network members. The thesis indicates that previous LCMs for calculating maintenance costs have often been very case-specific, suitable only for the item in question, and they have also been constructed for the needs of a single company, without the network perspective. The developed LCM is a proper tool for the decision making of maintenance services in the network environment; it enables analysing the past and making scenarios for the future, and offers choices between alternative maintenance operations. The LCM is also suitable for small companies in building active networks to offer outsourcing services for large companies. The research introduces also a five-step constructing process for designing a life-cycle costing model in the network environment. This five-step designing process defines model components and structure throughout the iteration and exploitation of user feedback. The same method can be followed to develop other models. The thesis contributes to the literature of value and value elements of maintenance services. It examines the value of maintenance services from the perspective of different maintenance network members and presents established value element lists for the customer and the service provider. These value element lists enable making value visible in the maintenance operations of a networked business. The LCM added with value thinking promotes the notion of maintenance from a “cost maker” towards a “value creator”.

The European Flood Alert System and the communication, perception, and use of ensemble predictions for operational flood risk management

Although ensemble prediction systems (EPS) are increasingly promoted as the scientific state-of-the-art for operational flood forecasting, the communication, perception, and use of the resulting alerts have received much less attention. Using a variety of qualitative research methods, including direct user feedback at training workshops, participant observation during site visits to 25 forecasting centres across Europe, and in-depth interviews with 69 forecasters, civil protection officials, and policy makers involved in operational flood risk management in 17 European countries, this article discusses the perception, communication, and use of European Flood Alert System (EFAS) alerts in operational flood management. In particular, this article describes how the design of EFAS alerts has evolved in response to user feedback and desires for a hydrographic-like way of visualizing EFAS outputs. It also documents a variety of forecaster perceptions about the value and skill of EFAS forecasts and the best way of using them to inform operational decision making. EFAS flood alerts were generally welcomed by flood forecasters as a sort of ‘pre-alert’ to spur greater internal vigilance. In most cases, however, they did not lead, by themselves, to further preparatory action or to earlier warnings to the public or emergency services. Their hesitancy to act in response to medium-term, probabilistic alerts highlights some wider institutional obstacles to the hopes in the research community that EPS will be readily embraced by operational forecasters and lead to immediate improvements in flood incident management. The EFAS experience offers lessons for other hydrological services seeking to implement EPS operationally for flood forecasting and warning. Copyright © 2012 John Wiley & Sons, Ltd.

Understanding climate data through commentary metadata: the CHARMe project

We describe the CHARMe project, which aims to link climate datasets with publications, user feedback and other items of "commentary metadata". The system will help users learn from previous community experience and select datasets that best suit their needs, as well as providing direct traceability between conclusions and the data that supported them. The project applies the principles of Linked Data and adopts the Open Annotation standard to record and publish commentary information. CHARMe contributes to the emerging landscape of "climate services", which will provide climate data and information to influence policy and decision-making. Although the project focuses on climate science, the technologies and concepts are very general and could be applied to other fields.

Elaboración de prototipos para el desarrollo de productos de tecnología social [

[ES] Concretamente, para el caso de las Matemáticas, encontramos numerosas aplicaciones orientadas al aprendizaje de operaciones básicas: suma, resta, multiplicación, división, pero ninguna que enseñe a los niños a aplicar dichas operaciones para la resolución de problemas.
 Y este es el fin de este Trabajo Fin de Grado; diseñar un prototipo de aplicación orientada a Tabletas, con tecnología Android, para la resolución de problemas matemáticos para alumnos de niños de 5º de primaria, con tres características fundamentales:
 1. La posibilidad de disponer de un espacio, en el cual los alumnos puedan realizar los cálculos necesarios para solucionar los problemas. 2. La posibilidad de incorporar un mecanismo, gracias al cual los usuarios puedan aportar nuevos problemas matemáticos, a modo de feedback con el usuario. 3. La combinación del aspecto lúdico con el académico. El principal objetivo de este Trabajo Fin de Grado es el de : desarrollar un prototipo orientado hacia dispositivos Tablet basados en Android, que permitan el aprendizaje sobre la materia de Matemáticas para un nivel de alumnos de 5º de primaria; y que cumpla con las características fundamentales anteriormente descritas.