No description available.

Thesis (M. Phil.)--University of Hong Kong, 1997. / Includes bibliographical references (leaf 221-248).

none

This thesis aims to contribute to the improvement of online machine translation software. We identify errors in the process of translation between English and French and make recommendations. The systems evaluated are Promt, Babylon, Google Translate and Bing and the reference corpus is taken from BankGloss. Promt made the most errors, followed by Babylon, Bing and Google. The systems together produced a total of 147 grammatical errors, 74 semantic errors, 17 lexical errors, and 6 stylistic errors. To improve Promt, we suggest expanding its dictionary. For Babylon, we advise adding more grammar rules. In order to reduce the number of semantic errors in Bing and Google, the software should learn to identify words according to context. Machine translation is not an end in itself, but a good aid in accomplishing translation tasks.

A presente pesquisa analisou a acurácia posicional horizontal das bases do Bing Maps, Google Maps e da World Imagery da Esri quando utilizadas como referência espacial on-line em um Sistema de Informação Geográfica no Município de São Paulo (MSP). A metodologia adotada foi a baseada no Decreto Federal no 89.817/84 e na Análise Estatística proposta por Merchant (1982). A análise da acurácia foi desenvolvida a partir das diferenças entre as coordenadas de 240 pontos nas cartas 1:1.000 do Mapa Digital da Cidade de São Paulo (MDC) em relação às coordenadas homólogas nas três bases, considerando separadamente as coordenadas do eixo Norte e Este. A base do Google Maps para o MSP foi dividida em duas (mosaico de ortofotos na área central e mosaico de imagens de satélite nas regiões periféricas), devido à grande diferença de acurácia entre os dois produtos. Para classificar cada base a partir do Decreto 89.817 foi definida a escala na qual somente 10% das discrepâncias tivessem seu valor superior ao PEC, e a escala na qual o Root Mean Square Error (RMSE) da amostra das discrepâncias fosse igual a 60,8% do PEC. A escala final selecionada foi a menor (menos detalhada) entre as definidas em cada um dos processos. A Análise Estatística foi baseada nos testes de tendência e precisão. Como as três bases apresentaram tendência, a escala definida pelo teste de precisão não foi considerada no cômputo final das escalas, devido à dificuldade de se eliminar a tendência nessas bases quando utilizadas no SIG. As escalas finais obtidas, relativas à classe A, foram: Google Maps (imagens de satélite): 1:12.400; Google Maps (ortofotos): 1:3.588; Bing Maps: 1:10.881 e Word Imagery da ESRI: 1:8.420. Concluiu-se que os três produtos com escalas próximas a 1:10.000 apresentam acurácia para serem utilizados como bases em SIGs nos estudos para planejamento urbano e que o Google Maps (ortofotos, com escala próxima a 1:4.000) pode ser igualmente utilizada para planejamento, mas em função de sua acurácia maior, pode servir também para a gestão de serviços urbanos. A principal limitação encontrada para as bases no uso como referência espacial em SIGs foi a inclinação das feições distantes do nadir da imagem ou da ortofoto e o consequente recobrimento de áreas adjacentes a essas feições. Entretanto, essa limitação se mostrou quase desprezível para as escalas definidas para as bases na análise da acurácia. / This research has analyzed the horizontal positional accuracy of basemaps Bing Maps, Google Maps and ESRIs World Imagery when used as an online spatial reference within a Geographic Information System for the municipality of São Paulo. The methodology was based on criteria defined by Brazil Federal Decree 89817/84 and in the analysis proposed by Merchant (1982). The accuracy analysis was developed observing the discrepancies between coordinates of selected 240 points from the 1:1000 digital map of São Paulo compared to corresponding points in the three basemaps, (coordinate directions North and East were considered separately). The Google Maps basemap for the city of São Paulo was divided in two (ortophoto mosaic for the central area and satellite images mosaic in the remainder peripheral areas), due to the considerable differences in their accuracy patterns. In order to classify each basemap as per Federal Decree 89.817, we have defined a scale in which only 10% of discrepancies were above the LMAS90 and the Root Mean Square Error (RMSE) of the discrepancies sample was equal to 60,8% of LMAS90. The final selected scale was the smallest (less detailed) of those obtained in each of the processes. The statistical analysis was based on the test of bias error and by a test of precision. Because the three basemaps have presented biases, the final scales defined by the precision test were not considered in the results, for it is difficult to eliminate biases in these basemaps when used in a GIS. We have obtained the following final scales to class A of the Brazilian Decree: Google Maps (area covered by satellite images): 1:12.400; Google Maps (area covered by ortophotos): 1:3.588; Bing Maps: 1:10.881 and ESRIs Word Imagery: 1:8.420. In conclusion, (a) the three products with scales around a 1:10.000 present accuracy to be used as basemap in GIS for urban planning studies and (b) Google Maps (area covered by ortophotos, scale around 1:4.000) can be equally used for planning studies, as well as urban services manager, due to its greater accuracy. The key limitations for the use of such basemaps as spatial references in GIS was the inclination of features which are distant from the image or ortophoto nadir (off-nadir effects) and the consequent shadowing of adjoining areas. However, this limitation is almost irrelevant to the scales defined for the basemaps in the accuracy analysis.

A presente pesquisa analisou a acurácia posicional horizontal das bases do Bing Maps, Google Maps e da World Imagery da Esri quando utilizadas como referência espacial on-line em um Sistema de Informação Geográfica no Município de São Paulo (MSP). A metodologia adotada foi a baseada no Decreto Federal no 89.817/84 e na Análise Estatística proposta por Merchant (1982). A análise da acurácia foi desenvolvida a partir das diferenças entre as coordenadas de 240 pontos nas cartas 1:1.000 do Mapa Digital da Cidade de São Paulo (MDC) em relação às coordenadas homólogas nas três bases, considerando separadamente as coordenadas do eixo Norte e Este. A base do Google Maps para o MSP foi dividida em duas (mosaico de ortofotos na área central e mosaico de imagens de satélite nas regiões periféricas), devido à grande diferença de acurácia entre os dois produtos. Para classificar cada base a partir do Decreto 89.817 foi definida a escala na qual somente 10% das discrepâncias tivessem seu valor superior ao PEC, e a escala na qual o Root Mean Square Error (RMSE) da amostra das discrepâncias fosse igual a 60,8% do PEC. A escala final selecionada foi a menor (menos detalhada) entre as definidas em cada um dos processos. A Análise Estatística foi baseada nos testes de tendência e precisão. Como as três bases apresentaram tendência, a escala definida pelo teste de precisão não foi considerada no cômputo final das escalas, devido à dificuldade de se eliminar a tendência nessas bases quando utilizadas no SIG. As escalas finais obtidas, relativas à classe A, foram: Google Maps (imagens de satélite): 1:12.400; Google Maps (ortofotos): 1:3.588; Bing Maps: 1:10.881 e Word Imagery da ESRI: 1:8.420. Concluiu-se que os três produtos com escalas próximas a 1:10.000 apresentam acurácia para serem utilizados como bases em SIGs nos estudos para planejamento urbano e que o Google Maps (ortofotos, com escala próxima a 1:4.000) pode ser igualmente utilizada para planejamento, mas em função de sua acurácia maior, pode servir também para a gestão de serviços urbanos. A principal limitação encontrada para as bases no uso como referência espacial em SIGs foi a inclinação das feições distantes do nadir da imagem ou da ortofoto e o consequente recobrimento de áreas adjacentes a essas feições. Entretanto, essa limitação se mostrou quase desprezível para as escalas definidas para as bases na análise da acurácia. / This research has analyzed the horizontal positional accuracy of basemaps Bing Maps, Google Maps and ESRIs World Imagery when used as an online spatial reference within a Geographic Information System for the municipality of São Paulo. The methodology was based on criteria defined by Brazil Federal Decree 89817/84 and in the analysis proposed by Merchant (1982). The accuracy analysis was developed observing the discrepancies between coordinates of selected 240 points from the 1:1000 digital map of São Paulo compared to corresponding points in the three basemaps, (coordinate directions North and East were considered separately). The Google Maps basemap for the city of São Paulo was divided in two (ortophoto mosaic for the central area and satellite images mosaic in the remainder peripheral areas), due to the considerable differences in their accuracy patterns. In order to classify each basemap as per Federal Decree 89.817, we have defined a scale in which only 10% of discrepancies were above the LMAS90 and the Root Mean Square Error (RMSE) of the discrepancies sample was equal to 60,8% of LMAS90. The final selected scale was the smallest (less detailed) of those obtained in each of the processes. The statistical analysis was based on the test of bias error and by a test of precision. Because the three basemaps have presented biases, the final scales defined by the precision test were not considered in the results, for it is difficult to eliminate biases in these basemaps when used in a GIS. We have obtained the following final scales to class A of the Brazilian Decree: Google Maps (area covered by satellite images): 1:12.400; Google Maps (area covered by ortophotos): 1:3.588; Bing Maps: 1:10.881 and ESRIs Word Imagery: 1:8.420. In conclusion, (a) the three products with scales around a 1:10.000 present accuracy to be used as basemap in GIS for urban planning studies and (b) Google Maps (area covered by ortophotos, scale around 1:4.000) can be equally used for planning studies, as well as urban services manager, due to its greater accuracy. The key limitations for the use of such basemaps as spatial references in GIS was the inclination of features which are distant from the image or ortophoto nadir (off-nadir effects) and the consequent shadowing of adjoining areas. However, this limitation is almost irrelevant to the scales defined for the basemaps in the accuracy analysis.

This thesis has been performed at consulting firm Sweco Position in Jönköping. The thesis is divided into two main sections: theoretical and practical. The theoretical section contains an investigation into the possibilities of implementing a geographical information system (GIS) for administrations of wells with the integration of Bing Maps for Windows 8 and Windows Phone 8. The practical work comprises the development of a prototype for Windows 8 and Windows Phone 8. The students review the Model-View-ViewModel-pattern on a theoretical basis, compares the differences between Windows 8 Pro and Windows RT. The thesis also contains and investigation into the pros and cons of a web based application compared to a native one. The result of the thesis will be comprised of a GIS-service developed in C# in co-operation with XAML with the options of saving data locally and a WMS-based service layer for Windows 8 and Windows Phone 8. / Detta examensarbete har utförts på konsultbolaget Sweco Position i Jönköping. Arbetet har varit uppdelat i två större delar: teoretiskt och praktiskt arbete. Teoretiska delen omfattade en utredning kring möjligheterna att implementera ett geografiskt informationssystem (GIS) för administration av brunnar med hjälp av Bing Maps tillsammans med Windows 8 och Windows Phone 8. Praktiska arbetet omfattade skapandet av en prototyp till Windows Phone 8 och Windows 8. Studenterna går bland annat genom Model-View-ViewModel-mönstret på en teoretisk nivå, beskriver skillnaderna mellan Windows 8 Pro och Windows RT. Ett utredande om för och nackdelar vid utveckling av webbaserade applikationer och native-applikationer. Resultatet av examensarbetet kommer innefatta en GIS-lösningen utvecklad i C# i samverkan med XAML och omfatta lokalt sparad data med WMS-tjänst baserad interaktion till Windows 8 och Windows Phone 8.

Thesis (M.CM)--Hong Kong Baptist University, 2006. / Dissertation submitted to the School of Chinese Medicine. Includes bibliographical references (leaves 24-26).

Thesis (M. CM)--Hong Kong Baptist University, 2006. / Dissertation submitted to the School of Chinese Medicine. Includes bibliographical references (leaves 18-19).

Thesis (M. CM)--Hong Kong Baptist University, 2006. / Dissertation submitted to the School of Chinese Medicine. Includes bibliographical references.