101 research outputs found
GeoQAMap - Geographic Question Answering with Maps Leveraging LLM and Open Knowledge Base (Short Paper)
No full textGeoQA (Geographic Question Answering) is an emerging research field in GIScience, aimed at answering geographic questions in natural language. However, developing systems that seamlessly integrate structured geospatial data with unstructured natural language queries remains challenging. Recent advancements in Large Language Models (LLMs) have facilitated the application of natural language processing in various tasks. To achieve this goal, this study introduces GeoQAMap, a system that first translates natural language questions into SPARQL queries, then retrieves geospatial information from Wikidata, and finally generates interactive maps as visual answers. The system exhibits great potential for integration with other geospatial data sources such as OpenStreetMap and CityGML, enabling complicated geographic question answering involving further spatial operations
Spatio-temporal weather maps: Extracting weather information along a travel route
Get PDFDenne masteroppgåva studerer romleg-temporale vêrkart. Meir spesifikt ser den på korleis ein kartbrukar kan få informasjon om vêret langs ei reiserute. Å få informasjon om vêret på ein enkel måte, når ein flytte seg i både tid og rom, er ei utfordrande oppgåve. Med bakgrunn i dette vart eit romleg-temporalt vêrkart laga i prosjektoppgåva i geomatikk, hausten 2022. Dette er eit kart som viser vêret på punkt langs ruta på det tidspunkt det er venta at ein vil passere punktet. Med bakgrunn i dette var det interesant å finne ut om det nye kartet virkeleg var betre enn klassiske vêrkart og vêrtabellar. For å finne ut av dette vart eit eksperiment utforma, programmert og utført. Resultatet av eksperimentet syner at det nye vêrkartet gjer det enklare for brukarane å få informasjonen dei treng om vêret langs reiseruta. Det finst også prov for at den nye metoden er raskare ein dei klassiske metodane. Derimot er det ikkje nokon ulikskap på metodane når målet er å finne ut når ein kjem fram til ulike punkt langs ruta. Enkelte av grunnane til ulikskap mellom på metodane er identifisert og diskutert. Nokre forslag til forbetringar som kan gjerast i framtida er også lagt fram.This thesis explore spatio-temporal maps for extracting information about the weather along a route. Easy comparison of weather information if there is change in space, in addition to time, is a challenging task. To address this problem, a spatio-temporal map was created in the geomatics specialization project in the fall of 2022. This new approach shows the weather at the presumed arrival time at intermediate points along the route. This facilitated the necessity to compare this new map with more classical weather maps and weather tables. An experiment was designed, programmed, and executed to try and figure this out. The results show that the new approach made it easier for the users to find the information they needed. There is also evidence that the map created in the specialization project is faster to use than the classical approaches. All three approaches do however perform the same if the goal is to find the arrival time at an intermediate location along the route. Some of the reasons for the differences between the methods are identified and discussed. There are also made suggestions for improvements that can be applied to the methods in the future
City Bike Usage Pattern Analysis and Demand Prediction Using Open Data
Get PDFUlike typer sykkeldelingssystemer ekspanderer globalt og er å finne i mange byer verden over. Sykkeldelingssystemer har vist tegn til å medføre positive effekter da de kan være alternativer til andre transportmidler. Studier har vist at bruk av sykkeldeling kan lede til positive helseeffekter og bidra i håndteringen av de pågående klimaendringene. Ekspansjonen av sykkeldelingssystemer fører med seg behov og vilje til å undersøke og forstå bruken av systemene. Flere studier har gjort nettopp dette ved å bruke generert bruksdata fra forskjellig sykkeldelingssystemer.
Arbeidet i denne oppgaven analyserer Trondheim Bysykkel, et stasjonsbasert sykkeldelingssystem i Trondheim, Norge. To oppgaver er utført basert på data gitt av Trondheim Bysykkel i tillegg til data fra flere eksterne kilder. En korrelasjonsanalyse er utført for tre forskjellige perioder, motivert av å forstå hvordan forskjellige faktorer påvirker bruken av systemet og identifisere viktige faktorer. Dette ble gjort ved hjelp av robust lineær regresjon, hvor fritidsfaktorer og høyden til stasjoner stod fram som innflytelsesrike. For at leverandørene av systemet skal kunne optimalisere den manuelle refordelingen av sykler mellom stasjoner, er det hensiktsmessig å predikere systembruken. Derfor er det i oppgaven implementert fire prediksjonsmodeller som skal predikere antall sykler levert og hentet ut hver time fra hver stasjon. Prediksjonene gjøres basert på en rekke ulike inputvariabler. Tre av de fire modellene predikerte relativt likt, med en gjennomsnittlig absolutt feil på omtrent 0.7 når det predikeres for alle stasjoner. I et eksperiment med fokus på de mest brukte stasjonene i hverdagsrushen, ble det predikert med en gjennomsnittlig absolutt feil rundt 1.3.Various types of bike-sharing systems are expanding globally and can be found in many cities world wide. Implementation of bike-sharing systems has shown signs of positive effects as they can be an alternative to other transportation modes. Studies have shown that usage of bike-sharing can lead to positive health effects and contribute to tackling ongoing climate changes.
With the expansion of bike-sharing comes need and desire to investigate and understand the dynamics of the usage. Several studies have done this over the last years by utilizing generated usage data of various bike-sharing systems.
The work of this thesis contributes with an investigation of Trondheim Bysykkel, a dock-based bike-sharing system in Trondheim, Norway. Based on data provided by Trondheim Bysykkel in combination with other data retrieved from several external sources, two main tasks are performed. Motivated by understanding how various factors influence the system usage and identify important features, a correlation analysis was conducted across three different periods. This was done through robust linear regression, where elevation of stations and leisure factors stood out as especially influential. Forecasting the usage demand is beneficial for service providers in order to optimize the process of manually redistribute bikes between stations. Hence, four prediction models were implemented to predict the hourly pick-up and drop-off demand at each station. The models predict based on a set of input features. Three of the four models achieved relatively similar performance, with a mean absolute error of approximately 0.7 when predicting for all stations. In a experiment focusing on the top most used stations during weekday rush hours, a mean absolute error of around 1.3 was achieved
Visuelle Analyse von großen Daten bewegender Autos – eine Überbrückung zwischen Thematischer Kartographie und Wissenschaftlicher Visualisierung
Get PDFThe aim of this thesis is to bridge the gaps between thematic mapping and scientific visualization and to achieve their synergetic effects for the visual analysis of big data. The author conducted a systematical comparative study of thematic cartography and scientific visualization. The results showed that these two disciplines reveal different visual analytical levels and are mutually complementary. Based on the theoretical findings, the author conducted extensive experiments of visually analyzing massive and complex real-world taxi floating car data. The results have confirmed our hypothesis that the two disciplines can achieve synergetic effects by taking advantage of their complementary characteristics.Das Ziel dieser Arbeit ist zwischen der Thematischen Kartographie und der Wissenschaftlichen Visualisierung eine Brücke zu bauen und die Synergieeffekte beider Disziplinen für die visuelle Analyse von Big Data zu gewinnen. Zu Bildung der theoretischen Grundlage findet eine systematische Vergleichsanalyse statt. Die Ergebnisse haben unterschiedliche visuell-analytische leistungen der beiden Disziplinen sowie deren komplementäre Rollen gezeigt. Die Autorin führt umfangreiche Untersuchungen der visuellen Analyse von Taxi-Bewegungsdaten durch. Die Ergebnisse haben die Ausgangshypothese bestätigt, dass die beiden Disziplinen aufgrund ihrer komplementären Eigenschaften vorteilhafte Synergieeffekte erzeugen können
An Ontology Based Prototype for Geocoding Offset Addresses
No full textThis paper explores the use of ontology based methods to determine the geographic locations of Chinese offset addresses Offset addresses are those addresses which contain a natural language location description reference to another location The study has two parts, 1) the definition of a geo-spatial ontology for Chinese addresses, and 2) the design of spatial computation rules to enable spatial reasoning and location computation. Web Ontology Language (OWL) is used in the first part to describe domain concepts, relationship between them. Semantic Web Rule Language (SWRL) is used in the second part to describe the rules that enable the process of spatial computation and reasoning The computed result after reasoning can be saved to the knowledge base and employed as new knowledge A prototype is built to verify, the model and the algorithm. The results demonstrate the feasibility of approach. They also revealed potential problems for spatial computations based on ontology.http://gateway.webofknowledge.com/gateway/Gateway.cgi?GWVersion=2&SrcApp=PARTNER_APP&SrcAuth=LinksAMR&KeyUT=WOS:000278560600048&DestLinkType=FullRecord&DestApp=ALL_WOS&UsrCustomerID=8e1609b174ce4e31116a60747a720701Computer Science, Software EngineeringEICPCI-S(ISTP)
Probabilistic Collocation Method for Yield Approach Index of Heterogeneous Slope
No full textAbstractA stochastic uncertainty quantification approach has been introduced into heterogeneous slope stability evaluation with the aid of probabilistic collocation method. This method combines Karhunen-Loeve expansion and polynomial chaos expansion to estimate the stochastic properties by running the deterministic geomechanics simulation models independently. With this approach, yield approach index of a heterogeneous slope is computed. The results show that in the toe and middle region FAI have the largest value indicating more instability, and the standard deviation field demonstrates that uncertainty has the largest value in the toe
- …
