Music emotion recognition (MER) on recommendation systems is becoming a topic of discussion in many new reasearches, e.g. on (Panda and Paiva 2011). In the musicrecommendationsystem described in this work, the recommendations that use emotional annotations rather than music similarity, produced better results for users that intend to listen to music according to a certain emotion or combination of emotions. Although this statement may be considered subjective, the results presented on figures 16A and 16B strengthen this idea, as the pair of recommendation strategies, that order songs by music annotation proximity on the V- A emotion model (R1 and R2), averaged higher classifications, than strategies applying psychoacoustic models, specifically considering timbral features obtained via MFCC extraction (R3 and R4). It is also important to keep in mind that these classifications take into account the emotional perception of users in a listening context, not the emotions that songs, individually or as a group, actually produce on the listener. Moreover, there’s no relevant difference between both recommendations based on emotional annotations and both recommendations that apply music similarity.
discovery and recommendation. Although, the final results point in that direction, after the experiments, 3 beta-testers stated that their music listening habits are not focused on the music artists they are listening to. Instead, they simply pay attention to the songs (mostly, the popular ones), and so, their playlists are track-driven, not artist-driven. That might have had presented a problem to those users, since the focus of RAMA is the relations between the artists. However, Spotify’s API’s recommendationsystem proved to please those users, who started to pay more attention to the name of the artists they listen to.
Last.FM was the first successful social network entirely dedicated to music, appearing in 2002 as a platform for music lovers. While Pandora was very big on the United States, being created in the United Kingdom, Last.FM started to expand its user base essentially in Europe. It was very important to the music dematerialization movement, however the radio only allowed non-paying users to listen to 30-seconds previews of the tracks. Last.FM distinguishes by its social features, with an automatically generated profile for every user, along with its main functionality, the internet streaming radio. Its huge success helped building a big user base, essential for the efficiency of the collaborative based recommendationsystem, proving that “better algorithms are nice but better data is nicer” (Krause, 2006). Behind the model used to build the similarity model is a method called audio-scrobbling, responsible for automatically logging all the songs played by the user. Last.FM also provides an optional plugin that monitors the user’s native media-player. This music library creates the conditions for the efficiency of the recommendationsystem, even without knowing about the songs’ inherent qualities. Although it uses tags associated to the tracks, it is based on the principle that if a user shares some favorite artists with a group of people, he will probably enjoy other popular artists within the group. Having essentially collaborative characteristics, it needs a certain amount of data to work with, taking some time to give recommendations feedback for new users. The Last.FM API allows developers to make use of the platform’s complex algorithms, getting the most popular albums of a certain artist or a list of similar artists to a given band or musician.
We have described the effort for deploying and evaluating a music recommender system on a real web site. The system is currently working online for registered users (free of cost) and responding in real time (www.palcoprincipal.pt). The recommender algorithm is a classical item-based collaborative filtering, extended with blacklists. Users are prompted with recommendation lists upon login and have the options of adding tracks to their playlists, adding tracks to their blacklists or ignoring recommendations. The impact of the activity on playlist addition has been measured using an A/B test, and a before/after analysis. Results strongly indicate an increase in playlist activity in the site generated by the recommender. The blacklist facility is also frequently used which shows the need for blacklists. The recommender system is implemented in SQL and is running since April 9th 2010. Response times are unnoticeable to the users. Model building time enables daily model refresh.
Abstract Literacy and music present important relations for the development of the child and this connection favors the acquisition of reading and writing. Through this partnership it is important to rethink strategies in the relationship of teaching and learning and to understand that literacy practices go beyond reading and writing activities per se. Music contributes to child neurodevelopment and naturally inserts the child into everyday social practices. This idea corroborates with recent discoveries of neuroscience and techniques of music therapy that indicate the effective contribution in the acquisition of reading and writing, through diverse practices in diverse modalities and styles .
de suas produções, deixando na história fonográfica deste gênero álbuns e gravações de questionável qualidade. Numa outra perspectiva, sua extensão efetiva aos dias atuais encontra-se diretamente relacionada ao próprio desenvolvimento do carnaval soteropolitano, e suas múl- tiplas atividades inter-relacionadas. Dentre elas, desta- que para os blocos de cordas, e o conjunto de organiza- ções empresariais advindos das estrelas e artistas deste segmento musical, motivando discussões e embates ideo- lógicos acerca de elementos presentes e constituintes de aspectos circunscritos a tradição e modernidade. Entretanto, ainda hoje, não raro, a constante presença e legitimação da Axé music no cenário musical local e nacio- nal é marcado por dissensões e mitos – estes, compreen- didos enquanto ideias não correspondentes com a verdade do fato social. Dentre os mitos, neste trabalho, destaque para o da monocultura, da suposta baixa qualidade técnica e de sua tão propagada crise/decadência/desaparecimento.
Fuelled by the 1880 celebrations of the tercentenary of Camões’s death and the Ultimatum of 1890, national consciousness increased in the last decades of the century; this also affected musical life, as a quest to identify and create a distinctly “Portuguese music” preoccupied composers, musicians and music critics. In 1884-85, Joaquim José Marques used the occasion of the Lisbon première of Augusto Machado’s opera Laureanne to publish an article series on “national melody” and “national music”. Appearing in the newly established musical periodical Amphion, they addressed state indifference to art music and its public institutions; as such they followed the denunciations, brought into currency by the Generation of 1870, of the decadence of Portuguese political, social and cultural institutions. Marques linked the “Portuguese national melody, so original and plangently modulated by our ancestors” to the libidinous lunduns and modinhas described by earlier foreign visitors and still subsisting, it was claimed, in brothels, taverns and prisons, performed to guitar accompaniment by fado singers and criminals, with erotic and vulgar lyrics and lascivious gestures. The sad loss of that unique and original Portuguese musical “character” (owing to the pernicious influence of operatic contrafacta) was also denounced by Marques in an extensive article published in the Revista de Estudos Livres in 1884-85. In keeping with the positivist and republican outlook of the Revista de Estudos Livres, Marques followed, in proclaiming the uniqueness of the “national song”, the literary, historical, and anthropological studies of the review’s main editor, Teófilo Braga. He adopted the ethnic nativism of that maître à penser, rejecting, for instance, for the folk-song “São João de Beja”, not only the theories of Arab and Muslim musical descent, but also the idea of affinity with traditional Spanish chants. By the same logic, the author suggested more generally that the “Portuguese melody” was derived from the pre-Roman Turdetani tribe, and coeval folk-songs, so it was asserted, had retained the Portuguese “original” musical “character”. Marques
Recommender systems (RS) are a branch of information systems that are widely used in many real-world setups and can be particularly common in e-commerce websites. Recommender systems aim to ﬁlter items from a large catalog that is intractable for humans to explore. The ﬁltering criterion employed is usually intended to ﬁt users’ preferences, interests, tastes, or needs. In contrast to other information ﬁltering sys- tems, recommender techniques are aimed to be proactive, enhancing discovering mech- anisms, as users do not have to explicitly state their information needs. According to Jannach and Adomavicius , the main purpose of recommenders is to suggest good items for a user, which in practice translates to two scenarios: (1) predicting user’s rating scores (rating prediction) or (2) ranking items according to the estimated user preferences (top-N recommendation). Balabanović and Shoham  proposed one of the ﬁrst categorizations of RS in the literature, which is composed of three classes, named collaborative ﬁltering (CF), content based (CB), and hybrid recommendation techniques.
The recommendationsystem of cultivars in Brazil was initiated in the late 1960s and early 1970s, when public and/ or private breeding institutions founded regional crop-specific commissions. The South Brazilian Comission of Wheat and Soybean was created, among others. These committees orga- nized experimental networks and established rules for the release of new cultivars, such as the number of locations and years a line should be tested in, the cultivars indicated as controls and specific criteria to determine the commercial release of new lines, i.e., a yield level based on the control cultivars, and other attributes such as wide adaptation and product quality. The ex- perimental networks formed a cooperative system of test trials of the new lines, involving public and private companies, which performed the tests without any remuneration for the service. This cooperative system consisted of institutions/companies interested in research, extension or trade of a particular crop. Consequently, not only the institutions/companies involved in the breeding of the crop participated in the cooperative system. At annual meetings, the results of the regional test trials were presented and discussed. The release or disposal of experimental lines was determined on a collegiate basis, where each participating institution was entitled to one vote. In the end, based on data obtained in experiments carried out by the different institutions and when the previously established criteria had been met, the system issued an approval for the commercial release and the cultivar was included in the list of those officially recommended.
Optical music recognition (OMR) is an important tool to recognize a scanned page of music sheet automat- ically, which has been applied to preserving music scores. In this paper, we propose a new OMR system to recognize the music symbols without segmentation. We present a new classiﬁer named combined neural net- work (CNN) that offers superior classiﬁcation capability. We conduct tests on ﬁfteen pages of music sheets, which are real and scanned images. The tests show that the proposed method constitutes an interesting contribution to OMR.
In order to evaluate the performance of the hybrid User Trust method, two main experiments have been conducted. The resulting top N user recommendation is used to find the most trusted neighbouring users within a social tagging system. From these neighbours, a set of recommended user is constructed. The hybrid User Trust (UserTrust) method is compared with the previous methods; the user-based collaborative filtering with Pearson Correlation Coefficient (PCC) (Resnick et al., 1994), Tidal Trust (TT) (Golbeck, 2006), UserRec (Zhou et al., 2010), tag-based Similarity Trust approach (ST) (Bhuiyan et al., 2010) and incorporation of social network information in CF (PCC-SN) (Liu and Lee, 2010).
ABSTRACT: The Diagnosis and Recommendation Integrated System (DRIS) allows the interpretation of results of leaf analysis through relationships among nutrients, instead of the absolute and isolated concentration of each one, as it is used by the criterion of sufficiency range. The objective was to evaluate three procedures of calculation of DRIS indices, and to verify the efficiency of DRIS as interpretation method for the results of Brachiaria decumbens (Signal grass). The study was developed with the results of six experiments carried out in a greenhouse at Piracicaba, SP, with nutrient solution. Concentrations of N, P, K, Ca, Mg, S, Cu, Fe, Mn, and Zn were used in the samples of recently expanded leaf laminae of the grass. The validation of the DRIS method used results from an experiment with nitrogen and sulfur rates applied to the same grass from the Mundo Novo farm, Brotas, SP. DRIS indices were calculated according to two criteria to choose the ratio order of nutrients (F value and R value) and three ways to calculate the nutrient functions (methods of Beaufils, Jones, and Elwali & Gascho). Nutritional Balance Index (NBI), calculated according to the generated norms, presented negative and significant correlation coefficients with the productivity in the combinations of methods tested and DRIS methods proposed by Beaufils, Jones and Elwali & Gascho were efficient in detecting concentrations that show nutrients deficiency or excess.
The Diagnosis and Recommendation Integrated System (DRIS) is a method to evaluate plant nutri- tional status that uses a comparison of the leaf tis- sue nutrient concentration ratios of nutrient pairs with norms from a high-yielding group (Soltanpour et al., 1995). The first step to implement DRIS or any other foliar diagnostic system is the establishment of standard values or norms (Walworth & Sumner, 1987; Bailey et al., 1997).
Due to the increasing offer of different types of multimedia content, consumers are becoming more demanding. At the same time, with the sheer size and diversity of online resources growing daily, it’s impossible for the users to analyse exhaustively all the available items to decide what is really relevant for them. Therefore, systems that are able to assist the user in selecting useful information, are becoming increasingly important towards meeting the users’ expectations. To address this problem, it is necessary to make use of a recommender system – the main area of this dissertation. Recommender systems (also called recommendation systems/platforms/engines) help the users to choose items they can find useful or of their interest. According to some early definitions, the main task of a recommender system is to “select some objects according to the user’s requirement” [ Wan98 ].
43 As revealed by previous research (BISCHOFF et al., 2008; HALPIN; ROBU; SHEPHERD, 2007), the distribution of tags by type varies along the power law curve. Therefore, with the aim of examining all tags and not only the most popular tags, we looked at the tags that had been assigned to the 184 songs in our sample. We randomly selected 181 tags (10 percent) from the 1,814 distinct combinations tag/song of the sample. The distribution is notably different from the one observed in the most popular tags (see Table 1 for details). The proportion of genre-related tags is lower at 40 percent. The analysis also shows a much higher proportion of tags in the Opinion (15 percent) and the Mood/Emotion (10 percent) categories. We also find a much higher number of tags (close to 9 percent) that did not fit in any of the main categories, most often because it was not possible to interpret what they meant (e.g., 332, AtS, MOONH8SUN). Finally, there is a significant proportion (more than 6 percent) of tags that consist in bibliographic information about a particular song, album, or artist (e.g., title of the song, artist name, music label), which, for obvious reasons, are not present in the most popular tags applied in Last.fm.
After watching all the recommended videos, there was an increased knowledge of the guidelines for water intake, the topic identified as the most in need of recommendation. This reinforces that the system can be useful to inform parents about lesser-known lifestyle guidelines. Strategies employed in this intervention, such as using an app with recommended videos, may have led to higher engagement of parents because all the parents in the intervention group watched their recommended contents. However, this may not be enough to make them aware that the children need to change their lifestyle because they first have to recognize the importance of changing the child’s weight status . This can also explain the low compliance with the guidelines despite their knowledge about them. The fact that this intervention was targeted to children with overweight and not only to children with obesity may make it difficult for parents to recognize their excessive weight. Studies indicated that the parents only recognized a child’s excessive weight if there was a substantial deviation in the child’s body size from perceived normality, especially if they were between the ages of 2 and 6 years old [37,38]. This indicates that new methodologies to motivate parents to manage their child’s weight have to be considered. One hypothesis would be to facilitate at least one session of motivational interviewing because it showed positive effects in the improvement of behaviors in parent–child health interventions, including those related to excessive weight . This may be included in an e-health format (such as a chat or a video consultation), as the development and testing of these e-health programs as a sole modality has been recommended .