894 resultados para Query expansion, Text mining, Information retrieval, Chinese IR
Resumo:
In this paper we study some of the characteristics of the art painting image color semantics. We analyze the color features of differ- ent artists and art movements. The analysis includes exploration of hue, saturation and luminance. We also use quartile’s analysis to obtain the dis- tribution of the dispersion of defined groups of paintings and measure the degree of purity for these groups. A special software system “Art Paint- ing Image Color Semantics” (APICSS) for image analysis and retrieval was created. The obtained result can be used for automatic classification of art paintings in image retrieval systems, where the indexing is based on color characteristics.
Resumo:
In recent years, learning word vector representations has attracted much interest in Natural Language Processing. Word representations or embeddings learned using unsupervised methods help addressing the problem of traditional bag-of-word approaches which fail to capture contextual semantics. In this paper we go beyond the vector representations at the word level and propose a novel framework that learns higher-level feature representations of n-grams, phrases and sentences using a deep neural network built from stacked Convolutional Restricted Boltzmann Machines (CRBMs). These representations have been shown to map syntactically and semantically related n-grams to closeby locations in the hidden feature space. We have experimented to additionally incorporate these higher-level features into supervised classifier training for two sentiment analysis tasks: subjectivity classification and sentiment classification. Our results have demonstrated the success of our proposed framework with 4% improvement in accuracy observed for subjectivity classification and improved the results achieved for sentiment classification over models trained without our higher level features.
Resumo:
In this paper we present algorithms which work on pairs of 0,1- matrices which multiply again a matrix of zero and one entries. When applied over a pair, the algorithms change the number of non-zero entries present in the matrices, meanwhile their product remains unchanged. We establish the conditions under which the number of 1s decreases. We recursively define as well pairs of matrices which product is a specific matrix and such that by applying on them these algorithms, we minimize the total number of non-zero entries present in both matrices. These matrices may be interpreted as solutions for a well known information retrieval problem, and in this case the number of 1 entries represent the complexity of the retrieve and information update operations.
Resumo:
In the context of Software Reuse providing techniques to support source code retrieval has been widely experimented. However, much effort is required in order to find how to match classical Information Retrieval and source code characteristics and implicit information. Introducing linguistic theories in the software development process, in terms of documentation standardization may produce significant benefits when applying Information Retrieval techniques. The goal of our research is to provide a tool to improve source code search and retrieval In order to achieve this goal we apply some linguistic rules to the development process.
Resumo:
2000 Mathematics Subject Classification: 62H30
Resumo:
Over the years there has been a broader definition of the term health. At the same time it was found also an evolution of the concept of health care which in turn has led to changes in the approach to delivery of health services and hence in its management. In this regard, currently the nephrology services have been searching for quality technical and social need. In view of these innovations and the quest for quality, it elaborated the general objective: to develop a quality assessment protocol for dialysis service Onofre Lopes University Hospital. It is an intervention project effected through an action research, which consisted of 4 steps. Initially was identified through a literature search in scientific literature, which quality indicators would apply to a dialysis unit being selected as follows: infection rate in hemodialysis access site, microbiological control of water used for hemodialysis and Index User satisfaction. Through critical reflection on the theme researched in the previous step, it was drawn up three data collection instruments, interview form type, applied between the months of October and November 2015. In addition to the information obtained, also made up of the use of information retrieval technique. The results were organized in graphs and tables and analyzed using qualitative and exploratory technical approach. Then a reflective analysis of the data obtained and the diagnosis of reality studied was traced and confronted with the literature was performed. The data produced in this study revealed that the Dialysis Unit of HUOL is much to be desired, considering that some weaknesses have been identified in its structure. Faced with this finding have been proposed, as a contribution and aiming to guide the development of future actions, suggestions for improvement that should be implemented and monitored to be assured overcoming these difficulties, allowing an appropriate organizational restructuring, and resulting in improved service public offered. It was concluded that for hemodialysis treatment results are achieved and positive, it is necessary to have physical structure and adequate infrastructure, multidisciplinary team specialized, trained and in sufficient quantity, well designed processes for professionals to have standards to be followed decreasing the chance to err, and a risk management system to detect and control situations that endanger patient safety.
Resumo:
Over the years there has been a broader definition of the term health. At the same time it was found also an evolution of the concept of health care which in turn has led to changes in the approach to delivery of health services and hence in its management. In this regard, currently the nephrology services have been searching for quality technical and social need. In view of these innovations and the quest for quality, it elaborated the general objective: to develop a quality assessment protocol for dialysis service Onofre Lopes University Hospital. It is an intervention project effected through an action research, which consisted of 4 steps. Initially was identified through a literature search in scientific literature, which quality indicators would apply to a dialysis unit being selected as follows: infection rate in hemodialysis access site, microbiological control of water used for hemodialysis and Index User satisfaction. Through critical reflection on the theme researched in the previous step, it was drawn up three data collection instruments, interview form type, applied between the months of October and November 2015. In addition to the information obtained, also made up of the use of information retrieval technique. The results were organized in graphs and tables and analyzed using qualitative and exploratory technical approach. Then a reflective analysis of the data obtained and the diagnosis of reality studied was traced and confronted with the literature was performed. The data produced in this study revealed that the Dialysis Unit of HUOL is much to be desired, considering that some weaknesses have been identified in its structure. Faced with this finding have been proposed, as a contribution and aiming to guide the development of future actions, suggestions for improvement that should be implemented and monitored to be assured overcoming these difficulties, allowing an appropriate organizational restructuring, and resulting in improved service public offered. It was concluded that for hemodialysis treatment results are achieved and positive, it is necessary to have physical structure and adequate infrastructure, multidisciplinary team specialized, trained and in sufficient quantity, well designed processes for professionals to have standards to be followed decreasing the chance to err, and a risk management system to detect and control situations that endanger patient safety.
Resumo:
MEDEIROS, Rildeci; MELO, Erica S. F.; NASCIMENTO, M. S. Hemeroteca digital temática: socialização da informação em cinema.In:SEMINÁRIO NACIONAL DE BIBLIOTECAS UNIVERSITÁRIAS,15.,2008,São Paulo. Anais eletrônicos... São Paulo:CRUESP,2008. Disponível em: http://www.sbu.unicamp.br/snbu2008/anais/site/pdfs/3018.pdf
Resumo:
This research addressed practice related problems from a medico-legal perspective and aims to provide a working tool that aids GPs to comply with best practice protocols. The resulting bag was developed in collaboration with General Practitioners, clinicians and members of the Medical Defense Union. Using proven methods developed within the Healthcare & Patient Safety Lab (e.g. DOME, Ambulance) to establish an evidence-based brief, this research used task, equipment and consumables analysis to determine minimum requirements and preferred layouts for task optimisation. The research established that clinicians require three distinct functions in their workspace: laying out, organisation and information retrieval. Feedback from clinicians indicates that this working tool allows them to access information and equipment wherever they may be and suggests an improvement from current practice. The research is now into a second year where the design of the bag will be refined and tested. Lifestyle and demographic changes such as the ageing population and increased prevalence of chronic diseases require more consistent standards of primary care, and care that is well coordinated and integrated (Imison, et al., 2011). Many guidelines exist relating to general practice and the doctor’s bag (NSLMC, 2008, RACGP, 2010, RCGP, 2008 and Hiramanek, 2004), however there is no standard in the UK that regulates the shape and materials of the bag or its contents. Doctors may use any sort of vessel to transport their equipment and consumables to a patient’s location. Furthermore, treating a patient in their own home, outside an ideal clinical environment, presents its own complications. A looks-like, works-like bag prototype and information system that will be used in clinical trials, the results of which will determine the manufacturing of a new, standardised bag for clinical treatment used by members of the Medical Defence Union.
Resumo:
A history of specialties in economics since the late 1950s is constructed on the basis of a large corpus of documents from economics journals. The production of this history relies on a combination of algorithmic methods that avoid subjective assessments of the boundaries of specialties: bibliographic coupling, automated community detection in dynamic networks, and text mining. These methods uncover a structuring of economics around recognizable specialties with some significant changes over the period covered (1956–2014). Among our results, especially noteworthy are (1) the clear-cut existence of ten families of specialties, (2) the disappearance in the late 1970s of a specialty focused on general economic theory, (3) the dispersal of the econometrics-centered specialty in the early 1990s and the ensuing importance of specific econometric methods for the identity of many specialties since the 1990s, and (4) the low level of specialization of individual economists throughout the period in contrast to physicists as early as the late 1960s.
Resumo:
MEDEIROS, Rildeci; MELO, Erica S. F.; NASCIMENTO, M. S. Hemeroteca digital temática: socialização da informação em cinema.In:SEMINÁRIO NACIONAL DE BIBLIOTECAS UNIVERSITÁRIAS,15.,2008,São Paulo. Anais eletrônicos... São Paulo:CRUESP,2008. Disponível em: http://www.sbu.unicamp.br/snbu2008/anais/site/pdfs/3018.pdf
Resumo:
As a way to gain greater insights into the operation of online communities, this dissertation applies automated text mining techniques to text-based communication to identify, describe and evaluate underlying social networks among online community members. The main thrust of the study is to automate the discovery of social ties that form between community members, using only the digital footprints left behind in their online forum postings. Currently, one of the most common but time consuming methods for discovering social ties between people is to ask questions about their perceived social ties. However, such a survey is difficult to collect due to the high investment in time associated with data collection and the sensitive nature of the types of questions that may be asked. To overcome these limitations, the dissertation presents a new, content-based method for automated discovery of social networks from threaded discussions, referred to as ‘name network’. As a case study, the proposed automated method is evaluated in the context of online learning communities. The results suggest that the proposed ‘name network’ method for collecting social network data is a viable alternative to costly and time-consuming collection of users’ data using surveys. The study also demonstrates how social networks produced by the ‘name network’ method can be used to study online classes and to look for evidence of collaborative learning in online learning communities. For example, educators can use name networks as a real time diagnostic tool to identify students who might need additional help or students who may provide such help to others. Future research will evaluate the usefulness of the ‘name network’ method in other types of online communities.
Resumo:
Institutions are widely regarded as important, even ultimate drivers of economic growth and performance. A recent mainstream of institutional economics has concentrated on the effect of persisting, often imprecisely measured institutions and on cataclysmic events as agents of noteworthy institutional change. As a consequence, institutional change without large-scale shocks has received little attention. In this dissertation I apply a complementary, quantitative-descriptive approach that relies on measures of actually enforced institutions to study institutional persistence and change over a long time period that is undisturbed by the typically studied cataclysmic events. By placing institutional change into the center of attention one can recognize different speeds of institutional innovation and the continuous coexistence of institutional persistence and change. Specifically, I combine text mining procedures, network analysis techniques and statistical approaches to study persistence and change in England’s common law over the Industrial Revolution (1700-1865). Based on the doctrine of precedent - a peculiarity of common law systems - I construct and analyze the apparently first citation network that reflects lawmaking in England. Most strikingly, I find large-scale change in the making of English common law around the turn of the 19th century - a period free from the typically studied cataclysmic events. Within a few decades a legal innovation process with low depreciation rates (1 to 2 percent) and strong past-persistence transitioned to a present-focused innovation process with significantly higher depreciation rates (4 to 6 percent) and weak past-persistence. Comparison with U.S. Supreme Court data reveals a similar U.S. transition towards the end of the 19th century. The English and U.S. transitions appear to have unfolded in a very specific manner: a new body of law arose during the transitions and developed in a self-referential manner while the existing body of law lost influence, but remained prominent. Additional findings suggest that Parliament doubled its influence on the making of case law within the first decades after the Glorious Revolution and that England’s legal rules manifested a high degree of long-term persistence. The latter allows for the possibility that the often-noted persistence of institutional outcomes derives from the actual persistence of institutions.
Resumo:
New developments in higher education and research are having their repercussions in dailylicencing practice. Examples are; demands for perpetual access usage of licensed content incourse packs or virtual research environments text mining open access to publications. Atthe Knowledge Exchange workshop on LicencingPractice, twenty Experts discussed how these newdevelopments could be incorporated in licencing. The workshop consisted of four presentations oncurrent developments in licencing followed by threeparallel breakout sessions on the topics open access,new developments and data and text mining. This led toa lively exchange of ideas. Especially the aspect of dataand text mining provided valuable insights in how thiscould be incorporated in licencing. The Knowledge Exchange Licensing expert group willwork on how to implement the model provisions discussed. Input from the workshop was collected for a workshop with publishers to take place in March 2012 and will include these provisions in their licences. The various suggestions will be also shared with other international organisations working inthis field.