922 resultados para top-N recommendation
Resumo:
Recommendation systems aim to help users make decisions more efficiently. The most widely used method in recommendation systems is collaborative filtering, of which, a critical step is to analyze a user's preferences and make recommendations of products or services based on similarity analysis with other users' ratings. However, collaborative filtering is less usable for recommendation facing the "cold start" problem, i.e. few comments being given to products or services. To tackle this problem, we propose an improved method that combines collaborative filtering and data classification. We use hotel recommendation data to test the proposed method. The accuracy of the recommendation is determined by the rankings. Evaluations regarding the accuracies of Top-3 and Top-10 recommendation lists using the 10-fold cross-validation method and ROC curves are conducted. The results show that the Top-3 hotel recommendation list proposed by the combined method has the superiority of the recommendation performance than the Top-10 list under the cold start condition in most of the times.
Resumo:
Dissertação de mestrado integrado em Engenharia Eletrónica Industrial e Computadores
Resumo:
With the rise of smart phones, lifelogging devices (e.g. Google Glass) and popularity of image sharing websites (e.g. Flickr), users are capturing and sharing every aspect of their life online producing a wealth of visual content. Of these uploaded images, the majority are poorly annotated or exist in complete semantic isolation making the process of building retrieval systems difficult as one must firstly understand the meaning of an image in order to retrieve it. To alleviate this problem, many image sharing websites offer manual annotation tools which allow the user to “tag” their photos, however, these techniques are laborious and as a result have been poorly adopted; Sigurbjörnsson and van Zwol (2008) showed that 64% of images uploaded to Flickr are annotated with < 4 tags. Due to this, an entire body of research has focused on the automatic annotation of images (Hanbury, 2008; Smeulders et al., 2000; Zhang et al., 2012a) where one attempts to bridge the semantic gap between an image’s appearance and meaning e.g. the objects present. Despite two decades of research the semantic gap still largely exists and as a result automatic annotation models often offer unsatisfactory performance for industrial implementation. Further, these techniques can only annotate what they see, thus ignoring the “bigger picture” surrounding an image (e.g. its location, the event, the people present etc). Much work has therefore focused on building photo tag recommendation (PTR) methods which aid the user in the annotation process by suggesting tags related to those already present. These works have mainly focused on computing relationships between tags based on historical images e.g. that NY and timessquare co-exist in many images and are therefore highly correlated. However, tags are inherently noisy, sparse and ill-defined often resulting in poor PTR accuracy e.g. does NY refer to New York or New Year? This thesis proposes the exploitation of an image’s context which, unlike textual evidences, is always present, in order to alleviate this ambiguity in the tag recommendation process. Specifically we exploit the “what, who, where, when and how” of the image capture process in order to complement textual evidences in various photo tag recommendation and retrieval scenarios. In part II, we combine text, content-based (e.g. # of faces present) and contextual (e.g. day-of-the-week taken) signals for tag recommendation purposes, achieving up to a 75% improvement to precision@5 in comparison to a text-only TF-IDF baseline. We then consider external knowledge sources (i.e. Wikipedia & Twitter) as an alternative to (slower moving) Flickr in order to build recommendation models on, showing that similar accuracy could be achieved on these faster moving, yet entirely textual, datasets. In part II, we also highlight the merits of diversifying tag recommendation lists before discussing at length various problems with existing automatic image annotation and photo tag recommendation evaluation collections. In part III, we propose three new image retrieval scenarios, namely “visual event summarisation”, “image popularity prediction” and “lifelog summarisation”. In the first scenario, we attempt to produce a rank of relevant and diverse images for various news events by (i) removing irrelevant images such memes and visual duplicates (ii) before semantically clustering images based on the tweets in which they were originally posted. Using this approach, we were able to achieve over 50% precision for images in the top 5 ranks. In the second retrieval scenario, we show that by combining contextual and content-based features from images, we are able to predict if it will become “popular” (or not) with 74% accuracy, using an SVM classifier. Finally, in chapter 9 we employ blur detection and perceptual-hash clustering in order to remove noisy images from lifelogs, before combining visual and geo-temporal signals in order to capture a user’s “key moments” within their day. We believe that the results of this thesis show an important step towards building effective image retrieval models when there lacks sufficient textual content (i.e. a cold start).
Resumo:
Moreira, A, Arsati, F, Cury, PR, Franciscon, C, Oliveira, PR, and Araujo, VC. Salivary immunoglobulin a response to a match in top-level brazilian soccer players. J Strength Cond Res 23(7): 1968-1973, 2009-It has been suggested that several parameters of mucosal immunity, including salivary immunoglobulin A (s-IgA), are affected by heavy exercise either in field sports or in the laboratory environment. Few observations have been made during a true sporting environment, particularly in professional soccer. We tested the hypothesis that salivary IgA levels will be decreased after a 70-minute regulation in a top-level professional soccer friendly match. Saliva samples from 24 male professional soccer players collected before and after the match were analyzed. Salivary immunoglobulin A concentration was measured by enzyme-linked immunosorbent assay and expressed as the absolute concentration (s-IgAabs), s-IgA relative to total protein concentration (IgA-Pro), and the secretion rate of IgA (s-IgArate). Rate of perceived exertion (RPE) was used to monitor the exercise intensity. The paired t-test showed no significant changes in s-IgAabs and s-IgArate (p > 0.05) from PRE to POST match. However, a significant (p < 0.05) increase in total protein concentration (1.46 +/- 0.4 to 2.00 +/- 07) and a decrease in IgA-Pro were observed. The best and most significant correlation was obtained with the RPE and changes in IgA-Pro (rs = -0.43) and could indicate that this expression may be an interesting marker of intensity in a soccer match. However, further investigation regarding exercise intensity, protein concentration, and immune suppression, particularly in team sports, is warranted. From a practical application, the variability of the responses among the players leads us to suggest that there is a need to individually analyze the results with team sports. Some athletes showed a decrease in s-IgA expressions, suggesting the need for taking protective actions to minimize contact with cold viruses or even reducing the training load.
Resumo:
We have tested the hypothesis that salivary cortisol increases after a competitive training match in top-level male professional soccer players divided in team A (n = 11) versus team B (n = 11). Saliva samples collected before and after the match were analyzed. Salivary cortisol concentrations were measured by enzyme-linked immunosorbent assay. The results from a two-way ANOVA with repeated measures showed no significant changes in salivary cortisol between either teams or time points (P > 0.05). Further investigation regarding competitive matches in a competition environment is warranted. In summary, the influence of intensive competitive training match alone appears to be minimal on salivary cortisol changes in top-level soccer adapted to this type of stress. From a practical application, the variability of the responses among the players leads us to suggest that there is a need to individually analyse the results with team sports.
Resumo:
This investigation examined the impact of a 17-d training period (that included basketball-specific training, sprints, intermittent running exercises, and weight training, prior to an international championship competition) on salivary immunoglobulin A (SIgA) levels in 10 subjects (athletes and staff members) from a national basketball team, as a biomarker for mucosal immune defence. Unstimulated saliva samples were collected at rest at the beginning of the preparation for the Pan American Games and 1 d before the first game. The recovery interval from the last bout of exercise was 4 h. The SIgA level was measured using enzyme-linked immunosorbent assay and expressed as absolute concentrations, secretion rate, and SIgA level relative to total protein. The decrease in SIgA levels following training was greater in athletes than in support staff; however, no significant differences between the two groups were detected. A decrease in SIgA level, regardless of the method used to express IgA results, was verified for athletes. Only one episode of upper respiratory tract illness symptoms was reported, and it was not associated with changes in SIgA levels. In summary, a situation of combined stress for an important championship was found to decrease the level of SIgA-mediated immune protection at the mucosal surface in team members, with greater changes observed in the athletes.
Resumo:
Moreira, A, Arsati, F, de Oliveira Lima-Arsati, YB, de Freitas, CG, and de Araujo, VC. Salivary immunoglobulin a responses in professional top-level futsal players. J Strength Cond Res 25(7): 1932-1936, 2011-The purpose of this study was to investigate the responses of salivary immunoglobulin A (SIgA) in 10 professional top-level Brazilian futsal players after 2 highly competitive games separated by 7 days. Unstimulated saliva was collected over a 5-minute period at PRE- and POST-match. The SIgA was measured by an enzyme-linked immunosorbent assay and expressed as the absolute concentration (SIgAabs) and secretion rate of IgA (SIgArate). Rate of perceived exertion and heart rate were used to monitor the exercise intensity. A 2-way analysis of variance with repeated measures showed nonsignificant differences between matches to SIgAabs, SIgArate, and saliva flow rate (p > 0.05). However, significant time differences were observed for all these parameters. In summary, we showed that a competitive training match induced a decrease in SIgA levels in top-level futsal players, which suggests an increment of the vulnerability to infections meditated by the training stimulus. This decrease suggests that the athletes were at an increased risk of developing an upper respiratory tract infection, and therefore, it could be necessary to take protective actions to minimize contact with cold viruses or even reduce the training load for athletes.
Resumo:
This paper presents the design of a low cost accessible digital television set-top box. This set-top box was designed and tested to the International ISDB-T system and considered the adoption of solutions that would provide accessible services in digital television in the simplest digital television receiver. The accessible set-top box was evaluated regarding the processing and memory requirements impacts to provide the features for accessible services. The work presents also the access services bandwidth consumption analysis(1).
Resumo:
Objective: To evaluate the frequency of overweight and obesity in health professionals, before and after a single specialized dietary recommendation. Methods: Anthropometric measures of 579 workers of a general hospital in the city of Sao Paulo, Brazil were taken. The weight (f), height (h) and waist circumference (wc) were interpreted according to the WHO and NCEP ATP III guidelines. Nutrition specialist provided dietary and behavioral recommendations. The entire sample underwent a new evaluation one year later. Results: At the first evaluation, 79 employees presente WC >= 102 cm (male) or WC >= 88 cm (female). The association between WC >= 102 cm (men) or WC >= 88 cm (women) and BMI >= 30 kg/m(2) was found in 12.8 % (69 subjects). The BMI distribution per age group indicated that the increase in overweight and obesity was directly proportional to the age increase. Physical activities were not practiced by 75% of the subjects studied. A year later, the evaluation indicated lack of statistical differences regarding the BMI and waist circumference of the sample and only 2.8% started to practice a physical activity. Conclusion: Dietary recommendation alone failed to promote changes in the eating habits of health professionals who work at a general hospital or to encourage them to practice exercise.
Resumo:
No contexto de reforma da administra????o p??blica, o objectivo do trabalho pretende dar conta das caracter??sticas dos modelos de direc????o p??blica, em particular dos dirigentes (Senior Executive Service: Top Public Service e Top Management Service), e do seu perfil em face dos novos desafios da administra????o p??blica. O objectivo principal deste trabalho foi caracterizar o sistema de designa????o/ nomea????o para os cargos de dirigente de livre escolha pol??tica, sistema conhecido na literatura como spoil system, com especial refer??ncia ao 1o e 2o n??veis, grupo I, dessas fun????es, no contexto da elite das direc????es-gerais em Portugal.
Resumo:
Fertilizer recommendation to most agricultural crops is based on response curves. Such curves are constructed from field experimental data, obtained for a particular condition and may not be reliable to be applied to other regions. The aim of this study was to develop a Lime and Fertilizer Recommendation System for Coconut Crop based on the nutritional balance. The System considers the expected productivity and plant nutrient use efficiency to estimate nutrient demand, and effective rooting layer, soil nutrient availability, as well as any other nutrient input to estimate the nutrient supply. Comparing the nutrient demand with the nutrient supply the System defines the nutrient balance. If the balance for a given nutrient is negative, lime and, or, fertilization is recommended. On the other hand, if the balance is positive, no lime or fertilizer is needed. For coconut trees, the fertilization regime is divided in three stages: fertilization at the planting spot, band fertilization and fertilization at the production phase. The data set for the development of the System for coconut trees was obtained from the literature. The recommendations generated by the System were compared to those derived from recommendation tables used for coconut crop in Brazil. The main differences between the two procedures were for the P rate applied in the planting hole, which was higher in the proposed System because the tables do not pay heed to the pit volume, whereas the N and K rates were lower. The crop demand for K is very high, and the rates recommended by the System are superior to the table recommendations for the formation and initial production stage. The fertilizer recommendations by the System are higher for the phase of coconut tree growth as compared to the production phase, because greater amount of biomass is produced in the first phase.
Resumo:
Alguns dos países da OCDE (2002) enfrentam, hoje, problemas relacionados ao emprego público e à sua competitividade (recrutamento, retenção de pessoal, falta crítica de habilidades e competências, etc.). Em nível de recrutamento (OCDE, 2002), os países que já identificaram o problema são o Canadá, a Dinamarca e a Finlândia; a falta crítica de habilidades e de competências foi já identificada por: Itália, Coreia, Noruega, Polónia, Portugal1 , Espanha, Áustria e Alemanha. O desenho e a implementação de estratégias de recursos humanos nas organizações públicas desenrolam-se em ambiente altamente politizado (LACOVIELLO, 1996) e o processo de decisão está sujeito a influências que tomam proporções maiores pelas características particulares que apresentam e pelo conflito essencial que decorre das relações entre políticos e burocratas.
Resumo:
The objective of this work was to evaluate the productive performance of broccoli under different top-dressing organic fertilizations. The experiment was conducted under protected cultivation, in a completely randomized design with four replications, with two plants per experimental unit. Broccoli seedlings were produced in a commercial substrate in styrofoam trays. The seedlings were transplanted to plastic pots containing 10.0 L of substrate made up of subsoil and organic compost at the ratio of 3:1 (v/v), respectively, which is equivalent to about 20.0 t ha-1 of organic compost at planting. After seedling establishment, the top-dressing fertilization treatments were applied: gliricidia biomass associated or not with liquid biofertilizer of cattle manure to the soil and bokashi. Two control treatments were established: one with mineral fertilization recommended for the crop and the other without top-dressing fertilization. The broccoli production was evaluated (commercial standard). Plants that received mineral fertilizer were more productive, however, they were not significantly different (p>0.05), by Dunnet test, from the plants fertilized with 2.5 t ha-1 gliricidiabiomass (dry mass) associated with liquid biofertilizer (2.0 L m-2) applied to soil. Top-dressing fertilizations with only gliricidia, at 2.5 and 5.0 t ha-1 of biomass (dry mass), resulted in no significant increase in production of broccoli inflorescence. The use of bokashi in addition to gliricidia biomass and liquid biofertilizer reduced the efficiency of the fertilization compared with plants that received only gliricidia and liquid biofertilizer.
Resumo:
This paper studies the use of results oriented performance information (outputs and outcomes) by local elected politicians (aldermen) in top managers’ performance evaluation. The main goal is to understand the top managers’ performance evaluation process conducted by Portuguese aldermen and to identify the dominant evaluation style. In line with Bogt (2001), an exploratory field research was developed in three large municipalities: Porto, Braga and Matosinhos. We intend to answer the following research questions: to what extent are local politicians (aldermen) 1 using results information (about outputs and outcomes) to evaluate top managers’ performance? Do aldermen give any importance to quantitative output information and report them on current planning and control documents? Will the output performance information be used on performance evaluation in a different way concerning policy field (task orientation)? Based on the Hopwood evaluation styles (adapted from the private to the public sector by Bogt, 2001), we have found that the performance evaluation process of top managers is more featured by the ‘operations-conscious styles’ (or ‘implicit style’). Portuguese aldermen recognize the importance to quantify outputs and outcomes, but these days they focus on the information about activities and the realized operations when they evaluate performance. We haven´t found significant differences between different tasks (some exceptions for culture and environment areas). However, the municipality of Matosinhos paid more attention on quantitative performance information. This finding is reinforced by the quality of quantitative data on outputs reported on its management report. Generally speaking, aldermen are aware that more attention should be paid to quantitative information about results and performance, taking into account recent reforms introduced in Portuguese municipalities (PLG) on this field, but the ‘implicit’ style is dominant.