Social Recommendation Based on Content and Trust

Report published in the Proceedings of the National Conference on "Education and Research in the Information Society", Plovdiv, May, 2014

A multimedia indexing and retrieval framework for multimedia database systems

The main challenges of multimedia data retrieval lie in the effective mapping between low-level features and high-level concepts, and in the individual users' subjective perceptions of multimedia content. ^ The objectives of this dissertation are to develop an integrated multimedia indexing and retrieval framework with the aim to bridge the gap between semantic concepts and low-level features. To achieve this goal, a set of core techniques have been developed, including image segmentation, content-based image retrieval, object tracking, video indexing, and video event detection. These core techniques are integrated in a systematic way to enable the semantic search for images/videos, and can be tailored to solve the problems in other multimedia related domains. In image retrieval, two new methods of bridging the semantic gap are proposed: (1) for general content-based image retrieval, a stochastic mechanism is utilized to enable the long-term learning of high-level concepts from a set of training data, such as user access frequencies and access patterns of images. (2) In addition to whole-image retrieval, a novel multiple instance learning framework is proposed for object-based image retrieval, by which a user is allowed to more effectively search for images that contain multiple objects of interest. An enhanced image segmentation algorithm is developed to extract the object information from images. This segmentation algorithm is further used in video indexing and retrieval, by which a robust video shot/scene segmentation method is developed based on low-level visual feature comparison, object tracking, and audio analysis. Based on shot boundaries, a novel data mining framework is further proposed to detect events in soccer videos, while fully utilizing the multi-modality features and object information obtained through video shot/scene detection. ^ Another contribution of this dissertation is the potential of the above techniques to be tailored and applied to other multimedia applications. This is demonstrated by their utilization in traffic video surveillance applications. The enhanced image segmentation algorithm, coupled with an adaptive background learning algorithm, improves the performance of vehicle identification. A sophisticated object tracking algorithm is proposed to track individual vehicles, while the spatial and temporal relationships of vehicle objects are modeled by an abstract semantic model. ^

Image retrieval on the Honeycomb Image Browser

Efficient and effective approaches of dealing with the vast amount of visual information available nowadays are highly sought after. This is particularly the case for image collections, both personal and commercial. Due to the magnitude of these ever expanding image repositories, annotation of all images images is infeasible, and search in such an image collection therefore becomes inherently difficult. Although content-based image retrieval techniques have shown much potential, such approaches also suffer from various problems making it difficult to adopt them in practice. In this paper, we follow a different approach, namely that of browsing image databases for image retrieval. In our Honeycomb Image Browser, large image databases are visualised on a hexagonal lattice with image thumbnails occupying hexagons. Arranged in a space filling manner, visually similar images are located close together enabling large image datasets to be navigated in a hierarchical manner. Various browsing tools are incorporated to allow for interactive exploration of the database. Experimental results confirm that our approach affords efficient image retrieval. © 2010 IEEE.

Special issue: Selected papers from the sixth IFIP 2.6 working conference on visual database systems: Guest editors' introduction

While multimedia data, image data in particular, is an integral part of most websites and web documents, our quest for information so far is still restricted to text based search. To explore the World Wide Web more effectively, especially its rich repository of truly multimedia information, we are facing a number of challenging problems. Firstly, we face the ambiguous and highly subjective nature of defining image semantics and similarity. Secondly, multimedia data could come from highly diversified sources, as a result of automatic image capturing and generation processes. Finally, multimedia information exists in decentralised sources over the Web, making it difficult to use conventional content-based image retrieval (CBIR) techniques for effective and efficient search. In this special issue, we present a collection of five papers on visual and multimedia information management and retrieval topics, addressing some aspects of these challenges. These papers have been selected from the conference proceedings (Kluwer Academic Publishers, ISBN: 1-4020- 7060-8) of the Sixth IFIP 2.6 Working Conference on Visual Database Systems (VDB6), held in Brisbane, Australia, on 29–31 May 2002.

Phenylketonuria: Protein content and amino acids profile of dishes for phenylketonuric patients. The relevance of phenylalanine

Phenylketonuria is an inborn error of metabolism, involving, in most cases, a deficient activity of phenylalanine hydroxylase. Neonatal diagnosis and a prompt special diet (low phenylalanine and natural-protein restricted diets) are essential to the treatment. The lack of data concerning phenylalanine contents of processed foodstuffs is an additional limitation for an already very restrictive diet. Our goals were to quantify protein (Kjeldahl method) and amino acid (18) content (HPLC/fluorescence) in 16 dishes specifically conceived for phenylketonuric patients, and compare the most relevant results with those of several international food composition databases. As might be expected, all the meals contained low protein levels (0.67–3.15 g/100 g) with the highest ones occurring in boiled rice and potatoes. These foods also contained the highest amounts of phenylalanine (158.51 and 62.65 mg/100 g, respectively). In contrast to the other amino acids, it was possible to predict phenylalanine content based on protein alone. Slight deviations were observed when comparing results with the different food composition databases.

Brokerage Platform for Media Content Recommendation

Near real time media content personalisation is nowadays a major challenge involving media content sources, distributors and viewers. This paper describes an approach to seamless recommendation, negotiation and transaction of personalised media content. It adopts an integrated view of the problem by proposing, on the business-to-business (B2B) side, a brokerage platform to negotiate the media items on behalf of the media content distributors and sources, providing viewers, on the business-to-consumer (B2C) side, with a personalised electronic programme guide (EPG) containing the set of recommended items after negotiation. In this setup, when a viewer connects, the distributor looks up and invites sources to negotiate the contents of the viewer personal EPG. The proposed multi-agent brokerage platform is structured in four layers, modelling the registration, service agreement, partner lookup, invitation as well as item recommendation, negotiation and transaction stages of the B2B processes. The recommendation service is a rule-based switch hybrid filter, including six collaborative and two content-based filters. The rule-based system selects, at runtime, the filter(s) to apply as well as the final set of recommendations to present. The filter selection is based on the data available, ranging from the history of items watched to the ratings and/or tags assigned to the items by the viewer. Additionally, this module implements (i) a novel item stereotype to represent newly arrived items, (ii) a standard user stereotype for new users, (iii) a novel passive user tag cloud stereotype for socially passive users, and (iv) a new content-based filter named the collinearity and proximity similarity (CPS). At the end of the paper, we present off-line results and a case study describing how the recommendation service works. The proposed system provides, to our knowledge, an excellent holistic solution to the problem of recommending multimedia contents.

A Multi–Agent Brokerage Platform for Media Content Recommendation

Near real time media content personalisation is nowadays a major challenge involving media content sources, distributors and viewers. This paper describes an approach to seamless recommendation, negotiation and transaction of personalised media content. It adopts an integrated view of the problem by proposing, on the business-to-business (B2B) side, a brokerage platform to negotiate the media items on behalf of the media content distributors and sources, providing viewers, on the business-to-consumer (B2C) side, with a personalised electronic programme guide (EPG) containing the set of recommended items after negotiation. In this setup, when a viewer connects, the distributor looks up and invites sources to negotiate the contents of the viewer personal EPG. The proposed multi-agent brokerage platform is structured in four layers, modelling the registration, service agreement, partner lookup, invitation as well as item recommendation, negotiation and transaction stages of the B2B processes. The recommendation service is a rule-based switch hybrid filter, including six collaborative and two content-based filters. The rule-based system selects, at runtime, the filter(s) to apply as well as the final set of recommendations to present. The filter selection is based on the data available, ranging from the history of items watched to the ratings and/or tags assigned to the items by the viewer. Additionally, this module implements (i) a novel item stereotype to represent newly arrived items, (ii) a standard user stereotype for new users, (iii) a novel passive user tag cloud stereotype for socially passive users, and (iv) a new content-based filter named the collinearity and proximity similarity (CPS). At the end of the paper, we present off-line results and a case study describing how the recommendation service works. The proposed system provides, to our knowledge, an excellent holistic solution to the problem of recommending multimedia contents.

Pesquisa de clip arts combinando imagens raster e vectoriais

Existe actualmente um crescente desenvolvimento de sistemas de armazenamento e pesquisa de imagens. Uma aproximação adoptada nesses sistemas é a recuperação de imagens baseada em conteúdo (CBIR, Content-Based Image Retrieval). No âmbito destas aplicações existem utilizadores que pretendem utilizar imagens clip art para os seus trabalhos e apresentações. Existem muitas imagens clip art espalhadas por diversas bases de dados em sítios na Internet ou em colecções vendidas em dispositivos ópticos. A pesquisa de imagens nestas bases de dados leva os utilizadores a percorrem várias listas de imagens manualmente ou por métodos de pesquisa por texto, muitas vezes ineficientes. Essas bases de dados de clip arts são representadas por imagens vectoriais e imagens raster. Existem várias tecnologias de pesquisa e recuperação de ambos os tipos de imagens clip art, raster e vectoriais, contudo, a investigação tem sido realizada em separado sem retirar partido das duas áreas de investigação em conjunto, no problema de recuperar e explorar colecções de clip arts. O objectivo deste trabalho é implementar um motor de busca para encontrar clip arts em base de dados compostas por imagens vectoriais e imagens raster. O trabalho envolve um conversor de imagens raster em vectoriais, a extracção de características das imagens raster e vectoriais e a avaliação do sistema de recuperação de clip arts.

Active shape structural model

Shape model, deformable models, structural models, biometry, content based image retrieval, sketches

Music identification

L'evolució ens els últims decennis de les possibilitats relacionades amb les tecnologies de la informació han provocat l'aparició de diferents camps, entre ells l'anomenat “recuperació de música basant-se en el contingut”, que tracta de calcular la similitud entre diferents sons. En aquest projecte hem fet una recerca sobre els diferents mètodes que existeixen avui en dia, i posteriorment n'hem comparat tres, un basat en característiques del so, un basat en la transformada discreta del cosinus, i un que combina els dos anteriors. Els resultats han mostrat, que el basat en la transformada de Fourier és el més fiable.

Food's visually perceived fat content affects discrimination speed in an orthogonal spatial task.

Choosing what to eat is a complex activity for humans. Determining a food's pleasantness requires us to combine information about what is available at a given time with knowledge of the food's palatability, texture, fat content, and other nutritional information. It has been suggested that humans may have an implicit knowledge of a food's fat content based on its appearance; Toepel et al. (Neuroimage 44:967-974, 2009) reported visual-evoked potential modulations after participants viewed images of high-energy, high-fat food (HF), as compared to viewing low-fat food (LF). In the present study, we investigated whether there are any immediate behavioural consequences of these modulations for human performance. HF, LF, or non-food (NF) images were used to exogenously direct participants' attention to either the left or the right. Next, participants made speeded elevation discrimination responses (up vs. down) to visual targets presented either above or below the midline (and at one of three stimulus onset asynchronies: 150, 300, or 450 ms). Participants responded significantly more rapidly following the presentation of a HF image than following the presentation of either LF or NF images, despite the fact that the identity of the images was entirely task-irrelevant. Similar results were found when comparing response speeds following images of high-carbohydrate (HC) food items to low-carbohydrate (LC) food items. These results support the view that people rapidly process (i.e. within a few hundred milliseconds) the fat/carbohydrate/energy value or, perhaps more generally, the pleasantness of food. Potentially as a result of HF/HC food items being more pleasant and thus having a higher incentive value, it seems as though seeing these foods results in a response readiness, or an overall alerting effect, in the human brain.

Trophic dynamics in the northern Humboldt Current system: insights from stable isotopes and stomach content analyses

The northern Humboldt Current system (NHCS) off Peru is one of the most productive world marine regions. It represents less than 0.1% of the world ocean surface but presently sustains about 10% of the world fish catch, with the Peruvian anchovy or anchoveta Engraulis ringens as emblematic fish resource. Compared with other eastern boundary upwelling systems, the higher fish productivity of the NHCS cannot be explained by a corresponding higher primary productivity. On another hand, the NHCS is the region where El Niño, and climate variability in general, is most notable. Also, surface oxygenated waters overlie an intense and extremely shallow Oxygen Minimum Zone (OMZ). In this context, the main objective of this study is to better understand the trophic flows in the NHCS using both stomach content and stable isotope analyses. The study focuses on a variety of organisms from low trophic levels such as zooplankton to top predators (seabirds and fur seals). The approach combines both long-term and specific studies on emblematic species such as anchoveta, and sardine Sardinops sagax and a more inclusive analysis considering the 'global' food web in the recent years (2008 – 2012) using stable isotope analysis. Revisiting anchovy and sardine we show that whereas phytoplankton largely dominated anchoveta and sardine diets in terms of numerical abundance, the carbon content of prey items indicated that zooplankton was by far the most important dietary component. Indeed for anchovy euphausiids contributed 67.5% of dietary carbon, followed by copepods (26.3%). Selecting the largest prey, the euphausiids, provide an energetic advantage for anchoveta in its ecosystem where oxygen depletion imposes strong metabolic constrain to pelagic fish. Sardine feed on smaller zooplankton than do anchoveta, with sardine diet consisting of smaller copepods and fewer euphausiids than anchoveta diet. Hence, trophic competition between sardine and anchovy in the northern Humboldt Current system is minimized by their partitioning of the zooplankton food resource based on prey size, as has been reported in other systems. These results suggest an ecological role for pelagic fish that challenges previous understanding of their position in the foodweb (zooplanktophagous instead of phytophagous), the functioning and the trophic models of the NHCS. Finally to obtain a more comprehensive vision of the relative trophic position of NHCS main components we used stable isotope analyses. For that purpose we analyzed the δ13C and δ15N stable isotope values of thirteen taxonomic categories collected off Peru from 2008 - 2011, i.e., zooplankton, fish, squids and air-breathing top predators. The δ15N isotope signature was strongly impacted by the species, the body length and the latitude. Along the Peruvian coast, the OMZ get more intense and shallow south of ~7.5ºS impacting the baseline nitrogen stable isotopes. Employing a linear mixed-effects modelling approach taking into account the latitudinal and body length effects, we provide a new vision of the relative trophic position of key ecosystem components. Also we confirm stomach content-based results on anchoveta Engraulis ringens and highlight the potential remarkable importance of an often neglected ecosystem component, the squat lobster Pleuroncodes monodon. Indeed, our results support the hypothesis according to which this species forage to some extent on fish eggs and larvae and can thus predate on the first life stages of exploited species. However, the δ13C values of these two species suggest that anchoveta and squat lobster do not exactly share the same habitat. This would potentially reduce some direct competition and/or predation.

QTL mapping for protein content in soybean cultivated in two tropical environments

The objectives of this study were to detect quantitative trait loci (QTL) for protein content in soybean grown in two distinct tropical environments and to build a genetic map for protein content. One hundred eighteen soybean recombinant inbred lines (RIL), obtained from a cross between cultivars BARC 8 and Garimpo, were used. The RIL were cultivated in two distinct Brazilian tropical environments: Cascavel county, in Paraná, and Viçosa county, in Minas Gerais (24º57'S, 53º27'W and 20º45'S, 42º52'W, respectively). Sixty-six SSR primer pairs and 65 RAPD primers were polymorphic and segregated at a 1:1 proportion. Thirty poorly saturated linkage groups were obtained, with 90 markers and 41 nonlinked markers. For the lines cultivated in Cascavel, three QTL were mapped in C2, E and N linkage groups, which explained 14.37, 10.31 and 7.34% of the phenotypic variation of protein content, respectively. For the lines cultivated in Viçosa, two QTL were mapped in linkage groups G and #1, which explained 9.51 and 7.34% of the phenotypic variation of protein content. Based on the mean of the two environments, two QTL were identified: one in the linkage group E (9.90%) and other in the group L (7.11%). In order for future studies to consistently detect QTL effects of different environments, genotypes with greater stability should be used.

Emotions in a Web-based Learning Environment

The aim of this thesis was to examine emotions in a web-based learning environment (WBLE). Theoretically, the thesis was grounded on the dimensional model of emotions. Four empirical studies were conducted. Study I focused on students’ anxiety and their self-efficacy in computer-using situations. Studies II and III examined the influence of experienced emotions on students’ collaborative visible and non-collaborative invisible activities and lurking in a WBLE. Study II also focused on the antecedents of the emotions students experience in a web-based learning environment. Study IV concentrated on clarifying the differences between emotions experienced in face-to-face and web-based collaborative learning. The results of these studies are reported in four original research articles published in scientific journals. The present studies demonstrate that emotions are important determinants of student behaviour in a web-based learning, and justify the conclusion that interactions on the web can and do have an emotional content. Based on the results of these empirical studies, it can be concluded that the emotions students experience during the web-based learning result mostly from the social interactions rather than from the technological context. The studies indicate that the technology itself is not the only antecedent of students’ emotional reactions in the collaborative web-based learning situations. However, the technology itself also exerted an influence on students’ behaviour. It was found that students’ computer anxiety was associated with their negative expectations of the consequences of using technology-based learning environments in their studies. Moreover, the results also indicated that student behaviours in a WBLE can be divided into three partially overlapping classes: i) collaborative visible ii) non-collaborative invisible activities, and iii) lurking. What is more, students’ emotions experienced during the web-based learning affected how actively they participated in such activities in the environment. Especially lurkers, i.e. students who seldom participated in discussions but frequently visited the online environment, experienced more negatively valenced emotions during the courses than did the other students. This result indicates that such negatively toned emotional experiences can make the lurking individuals less eager to participate in other WBLE courses in the future. Therefore, future research should also focus more precisely on the reasons that cause individuals to lurk in online learning groups, and the development of learning tasks that do not encourage or permit lurking or inactivity. Finally, the results from the study comparing emotional reactions in web-based and face-to-face collaborative learning indicated that the learning by means of web-based communication resulted in more affective reactivity when compared to learning in a face-to-face situation. The results imply that the students in the web-based learning group experienced more intense emotions than the students in the face-to-face learning group.The interpretations of this result are that the lack of means for expressing emotional reactions and perceiving others’ emotions increased the affectivity in the web-based learning groups. Such increased affective reactivity could, for example, debilitate individual’s learning performance, especially in complex learning tasks. Therefore, it is recommended that in the future more studies should be focused on the possibilities to express emotions in a text-based web environment to ensure better means for communicating emotions, and subsequently, possibly decrease the high level of affectivity. However, we do not yet know whether the use of means for communicating emotional expressions via the web (for example, “smileys” or “emoticons”) would be beneficial or disadvantageous in formal learning situations. Therefore, future studies should also focus on assessing how the use of such symbols as a means for expressing emotions in a text-based web environment would affect students’ and teachers’ behaviour and emotional state in web-based learning environments.

Recuperação de imagem baseada em conteúdo: uso de atributos de textura para caracterização de microcalcificações mamográficas

Resumo Este trabalho descreve um sistema de recuperação de imagens baseada em conteúdo ("content-based image retrieval"), desenvolvido para auxiliar o diagnóstico de lesões de mama por inspeção visual, mediante comparação de imagens. Quando uma imagem desconhecida é apresentada, o sistema extrai um vetor de atributos de textura e busca, em um banco de dados, imagens com características semelhantes dentro de uma aproximação previamente estabelecida. As imagens recuperadas são apresentadas ao usuário, que pode, então, verificar os diagnósticos associados.