2013 5th Computer Science and Electronic Engineering Conference (CEEC)
University of Essex, UK
Towards an Efficient Routing Web Processing Service through Capturing Real-time Road Conditions from Big Data Mohamed Bakillah
Amin Mobasheri
GIScience research group, Heidelberg University Heidelberg, Germany and Department of Geomatics Engineering University of Calgary, Alberta, Canada
[email protected]
GIScience research group University of Heidelberg Heidelberg, Germany
[email protected]
Steve H.L. Liang
Alexander Zipf
Department of Geomatics Engineering University of Calgary Alberta, Canada
[email protected]
GIScience research group University of Heidelberg Heidelberg, Germany
[email protected]
experimental platform to study the VGI phenomena and demonstrate all the opportunities for several applications. With these new promises, users are expecting that not only they will have access to large data sets, but more importantly, they will be able to pose more complex queries and infer more information than ever. However, research still need to be conducted to achieve these expectations, including with respect to information retrieval techniques. In this paper, we specifically focus on information retrieval for routing and navigation services. Geospatial information retrieval is integral part of routing and navigation services, notably to help find the best route based on several social events that can cause an obstacle and increase transportation costs. Several approaches for retrieving landmarks or points of interest are able to process queries to retrieve entities that exist in the source, such as stadium, hospital, lake, etc. [2]. However, existing approaches still have difficulties to resolve problems that are caused by the dynamic events taking place in reality. In fact, by taking into account the real-time dynamic events a better solution for routing services can be provided. In this paper, we propose our new event service architecture to improve the normal routing services by employing VGI as Big Data. The principle behind the approach is that spatial relations between entities stored in the database can reveal the existence of other implicit entities that are not stored in the database. In this case, we can show that the appropriate semantic modeling of these spatial relations can help to infer the existence of implicit entities. We have designed a web processing service architecture based on Zoo platform for event processing in support to routing services. The paper is organized as follows: in the next section, we present examining the latest developments and issues associated with big data from the perspective of the analysis of VGI. In Section 3, we present the routing service and elaborate on its current problems that need to be solved. In
Abstract— The rapidly growing number of crowdsourcing platforms generates huge volumes of volunteered geographic information (VGI), which requires analysis to reveal their potential. The huge volumes of data appear as an opportunity to improve various applications, including routing and navigation services. How existing techniques for dealing with Big Data could be useful for the analysis of VGI remains an open question, since VGI differs from traditional data. In this paper, we focus on examining the latest developments and issues associated with big data from the perspective of the analysis of VGI. This paper notably presents our new architecture for exploiting Big VGI in event service processing in support to optimization of routing service. In addition, our study highlights the opportunities that are created by the emergence of Big VGI and crowdsourced data on improving routing and navigation services, as well as the challenges that remain to be addressed to make this a reality. Finally, avenues for future research on the next generation of collaborative routing and navigation services are presented. Keywords- Big Data; VGI; routing service; event service
I.
INTRODUCTION
Sound decision-making in the geographical domain involves answering to complex queries, with inferring facts from available geospatial data sources. Meanwhile, the amount of available data has been rapidly growing, due, among other phenomena, to the increasing dissemination of digital sensors, smart phones, crowdsourcing applications, social media, etc. This gave rise to the concept of “Big Data,” i.e., volumes of data that exceed the processing capacity of conventional data systems [1]. The phenomenon of crowdsourcing in general, and volunteered geographic information (VGI) in particular, has contributed significantly to the phenomenon of Big Data. A well-known example is Open Street Map (OSM), which has now become an
978-1-4799-0383-2/13/$31.00 ©2013 IEEE
152
2013 5th Computer Science and Electronic Engineering Conference (CEEC)
novelty, they claim, lies in part in the community-based aspect of the users’ contribution to this digital commons of geographic knowledge [8]: VGI is often created out of the collaborative involvement of large communities of users in a common project – for example Open Street Map (OSM) or Wikimapia –, where individuals can produce geographic information that emanates from their own local knowledge of a geographic reality or edit information provided by other individuals. For example, in OSM, users can describe map features – such as roads, water bodies, and points of interest – using “tags,” providing information at a level of detail that often goes beyond the level of detail that can be provided by traditional geospatial data producers [9]. As a result, and with the ever increasing number of crowdsourcing applications, the volume of VGI is becoming huge, with no doubt that VGI is now an important component of Big Data. Among the advantages associated with VGI, researchers highlight its use to enrich, update, or complete existing geospatial data sets [10, 11, 12, 13]. This advantage is especially put forward in the context where traditional geospatial data producers – which are usually governments – may lack the capacity to generate data sets with comprehensive spatial and temporal coverage and level of detail [11, 12] such as needed for routing services to be efficient. Furthermore, it was highlighted that VGI can be provided and disseminated in a timely, near real-time fashion, which is highly required in routing services. The advantages associated with VGI strongly suggest that this type of knowledge is highly valuable and is likely to help providing a dynamic picture of the environment [14]. Nevertheless, the use of VGI for mobile applications such as routing and navigation is not yet fully achievable, as it is hampered by various obstacles related both to large volumes, heterogeneity, and credibility.
Section 4, the idea of employing an event processing service that processes Big VGI in order to overcome the problems of routing service is discussed. Section 5 presents the architecture of the system as well as details of the approach. Finally, conclusions and future work are provided in Section 6. II.
VGI AS BIG DATA
Recent technological advances in data production and dissemination have enabled the generation of unprecedented volumes of geospatial data. In 2012, it was estimated that the global volume of data was growing at a 50 percent rate each year [3], due, among other phenomena, to the increasing dissemination of digital sensors, smart phones, crowdsourcing applications, social media, etc. While geospatial data have traditionally remained at the hand of experts (governments, mapping agencies), paradigms such as open data, social media and collaborative mapping projects make it possible for an increasing proportion of this data to be available to virtually anyone, with the potential to benefit businesses, civil society and individuals in general. This huge amount of data has given rise to the term “Big Data.” Big Data is a loosely-used term that is employed to refer to two ideas. First, it is used to refer to this huge volume of data itself. For example, [1] writes that “Big Data is data that exceeds the processing capacity of conventional data systems” and does not “fit the structure of your database architectures,” because of its size, but also because of its dynamicity. Secondly, the term is also used, perhaps less frequently, to refer to the set of techniques that are being developed to deal with such volumes of data. For example, [4] reports that the term is used to refer to “smarter, more insightful analysis” of large volumes of data, while Oracle defines it as “techniques and technologies that enable enterprises to effectively and economically analyze all of their data” [5]. Indeed, Big Data in itself is of no great value unless we find means of managing and analysing this less conventional data, which is not necessarily formatted according to the usual rows and columns of traditional databases. The question raised by Big Data is therefore how to extract information and knowledge from these raw data streams, since traditional approaches are not suitable for such amount and heterogeneity of data coming from various sources [4,6]. Nowadays, users can produce geographic information via a variety of Internet applications; as a result, a “global digital commons of geographic knowledge” is created without having to rely solely on “traditional” geospatial data production processes [7]. In 2007, Goodchild introduced the term “volunteered geographic information” to refer to the geographic information generated by users through Web 2.0 era applications. Later, [2] stated that the VGI paradigm reflects the transformation of users from “passive” geospatial information consumers to “active contributors.” However, [8] argue that the concept of “user-generated content” is not new, referring for instance to public participation GIS where users can provide input and feedback to decision-makers and involved communities through Web-based applications. The
978-1-4799-0383-2/13/$31.00 ©2013 IEEE
University of Essex, UK
III.
ROUTING SERVICE
Routing of vehicles is nowadays a typical use of digital geographic data. One of the most popular services is the Google’s routing service, which was launched in 2005 (http://googleblog.blogspot.de/2005/02/mapping-yourway.html). Since then, the service has been improved by the addition of public transport data and the integration of realtime traffic information which started in 2008. By integrating new algorithms, especially from the field of hierarchical routing, worldwide routers improved their speed for long route calculation [15]. Especially in car routing issues, most of the static routing problems are solved. In terms of standardizing and interoperability, interfaces for locationbased services (LBS), and in particular for routing, were developed [16]. With the arrival of VGI and, notably, of the Open Street Map (OSM) project, a second generation of route planning services is starting to emerge. VGI applications make routable data more easily available for free [17] and reduce information gaps. For example, the OpenRouteService.org uses OSM as a source of data (Figure. 1). Users can add points-of-interest (POIs) as identified in OSM or search for POIs by name. Increasing the sources of data on POIs with
153
2013 5th Computer Science and Electronic Engineering Conference (CEEC)
other sources than OSM would allow to take into account more POIs but would require extensive integration work to merge different heterogeneous data sets.
best routing choice for a given source and destination. Therefore, the proposed event service analyses the data captured by sensors and via a reasoning service, considering the constraints and properties of a certain event derives additional information about the situation in a given region. This extra information is then passed as real-time information to the routing service in order to optimize the results. For example, for a “snow” event, the constraints of minimum and maximum level of snow can be defined in the event service. In addition, several rules can be defined, for instance, if the level of snow captured by a sensor is between the defined values for minimum and maximum thresholds, then a certain speed limit for vehicles should be considered. This speed limit value as well as the coordinates of the sensor is passed to the spatial processing service in order to estimate the time and costs of traveling within that area. Finally, the routing service can benefit from this information in order to give proper navigation information to users.
Figure 1. OpenRouteService.org with selection of POIs
In addition, special routing services such as wheelchair routing, bike routing, agricultural routing, to name only a few, are being designed. New approaches and questions came up, specifically in the field of data linking and using crowdsourced and sensor information. With the help of this new information it will be possible considering traffic load, weather conditions, and physical road conditions to support real-time route planning. IV.
V.
THE SYSTEM ARCHITECTURE
The Event service architecture for real-time routing service is presented in Figure 2. The main components of this architecture are the main services itself. Besides routing service, a Sensor Observation Service is employed to feed real-time data (in our example meteorological data) to the Event Processing Service. The SOS itself could be recognized by checking the metadata of service which could be provided via OGC Catalogue Service. Besides Sensor Data, the Event WPS exploits VGI (e.g. OSM), as well as the information about various event patterns, event constraints, rules, etc. provided by knowledgebase. In the next step, Event WPS checks the observed sensor data against the event patterns in order to perform reasoning and generate implicit information to update the existing road network. The updated network along with the event information is passed to the routing service which can be further used for optimal route finding and decision making.
OPTIMIZED ROUTING SERVICE USING EVENT SERVICE AND BIG VGI
Analyzing raw data from such variety of sources poses challenges due to high volume and heterogeneity of Big VGI and crowdsourced data. Traditional routing services can be optimized by using the additional information extracted from Big VGI. This could be done by a reasoning service that can analyze the crowdsourced as well as sensor data and understand spatio-temporal patterns occurred in a specific region. This extra information could later be used by the routing service for giving more accurate results based on real-time situation. In this paper, we aim to design an event spatial web processing service that exploits VGI as one of its inputs, and through a reasoning service, generates useful information for the routing service. The vast majority of today's event processing systems focus on the efficiency of reasoning algorithms. However, these don't take into account the various types of uncertainty that exist in most applications [18]. As big data applications, many of the emerging event processing systems are required to process events that arrive from sources such as sensors and volunteered sources, which have inherent uncertainties associated with them. In these cases, the streams of events may be incomplete or inaccurate, for example, regarding the time and location of events. Many state-of-the-art systems assume that data is precise and certain, or that it has been cleansed before processing. In this paper, we propose an event spatial processing service that exploits Open Street Map (OSM) data as well as meteorological data captured by sensors. It is believed that the weather condition plays a major role in examining the
978-1-4799-0383-2/13/$31.00 ©2013 IEEE
University of Essex, UK
Figure 2. The System architecture
154
2013 5th Computer Science and Electronic Engineering Conference (CEEC)
Note that, the architecture also includes a population estimation WPS which is able to monitor the population density for a given region in order to estimate traffic flow. The current routing service makes use of such a web service, and therefore this service along with our proposed Event WPS (or any other possible relevant services in SDI) could be discovered by the OGC discovering service and aggregated via the Service composition component. Finally, the result of all aggregated services/data are fed to the routing service which is in charge of accepting queries from users (e.g. humans, services) via a User Interface. The results of optimal route choices are then returned back to the user in a proper manner.
[2]
[3] [4] [5]
[6] [7] [8]
VI.
CONCLUSION AND FUTURE WORK
In this paper we presented an Event Service architecture for using Big VGI in order to prepare real-time routing service. We introduced the concept of VGI as Big Data, and emphasized on the usefulness of existing techniques for analyzing Big Data for VGI. Later we discussed the existing routing service, its capabilities, as well as the problems for existing solution for being real-time. Finally, in order to overcome this problem we presented the system architecture which employs an Event Spatial WPS and elaborated the configurations within such architecture. For future research the aim is to implement and test Event WPS with real data collected from sensors. In another path, the aim is to combine this event service architecture with other services/architecture like MATSIM and use it for improving evacuation simulation. The second class of research issues are supporting complex events and extending our Event WPS to support other applications such as disaster response, and increasing the explicitness power of the reasoning tools. Last but not least, since our service employs volunteered geographic data and the results heavily relies on the input data, the quality of VGI is an important aspect that should be considered. In this research we assume that data has an acceptable level of quality, however, in reality several research studies are working on the quality of geo-data such as [19, 20]. In future, the web processing service presented in this study could be coupled with a quality evaluation web service [20] in order to assure the quality of input data that is used for real-time routing service.
[9]
[10] [11]
[12] [13]
[14]
[15]
[16]
[17] [18]
ACKNOWLEDGMENT This Research was made possible with the funding support of Microsoft Research and Alberta Innovate Technology Future.
[19]
[20]
REFERENCES [1]
Dumbill, E. 2013. Making Sense of Big Data. Big Data 1(1).
978-1-4799-0383-2/13/$31.00 ©2013 IEEE
155
University of Essex, UK
Ballatore, A., Bertolotto, M. 2011. Semantically enriching VGI in support of implicit feedback analysis. In: K. Tanaka, P. Fröhlich, and K.-S. Kim (Eds.), Proc. W2GIS 2011, LNCS 6574, Springer-Verlag, Berlin Heidelberg, pp. 78-93. Lohr, Steve. 11 February 2012. The Age of Big Data. The New York Times. Davenport, T.H., Barth, P., Bean, R. 2012. How “Big Data” is Different. MIT Sloan Managament Review 54(1), 22-24. Oracle. 2012. Big Data Technical Overview. available at: http://www.oracle.com/in/corporate/events/bigdata-technicaloverview-pune-1902240-en-in.pdf (Accessed 30 March 2013) Birkin, M. 2012. Big Data Challenges for Geoinformatics. Geoinformatics & Geostatistics: An Overview 1(1). Hardy, D. 2010. Volunteered Geographic Information in Wikipedia. PhD Thesis, University of California, Santa Barbara, USA, 260 p. Coleman, D., Georgiadou, Y., Labonté, J. 2009. Volunteered geographic information: the nature and motivation of produsers. International Journal of Spatial Data Infrastructures Research, Special Issue GSDI-11. Goetz, M., Lauer, J., Auer, M. 2012. An algorithm-based methodology for the creation of a regularly updated global online map derived from volunteered geographic information. In: C.-P. Rückemann and B. Resch (Eds.), Proc. Fourth International Conference on Advanced Geographic Information Systems, Applications, and Services, Valencia, Spain, pp. 50-58. Goodchild, M.F. 2007. Citizens as sensors: the world of volunteered geography. GeoJournal 69(4), 211-221. Gupta, R. 2007. Mapping the Global Energy System Using Wikis, Open Sources, WWW, and Google Earth. Position paper presented at the Workshop on Volunteered Geographic Information, Santa Barbara, CA, December 13-14 2007, 2 p. Tulloch, D. 2008. Is volunteered geographic information participation? GeoJournal 72(3). Goetz, M., Zipf, A. 2011. Towards defining a framework for the automatic derivation of 3D CityGML models from volunteered geographic information. Joint ISPRS Workshop on 3D City Modelling & Applications and the 6th 3D GeoInfo Conference, Wuhan, China. Mooney, P., Corcoran, P. 2011. Annotating spatial features in OpenStreetMap. Proc. GISRUK 2011, Portsmouth, England, April 2011. Delling, D., Sanders, P., Schultes, D., Wagner, D. 2009. Engineering Route Planning Algorithms. J. Lerner, D. Wagner, & K. A. Zweig, (Eds.) Algorithmics of large and complex networks, 2, 117–139. doi:10.1007/978-3-540-72845-0_2. Mabrouk, M., Bychowski, T., Williams, J., Niedzwiadek, H., Bishr, Y., Gaillet, J.-F., Crisp, N., et al. 2005. OpenGIS Location Services (OpenLS): Core Services, OpenGIS® Im- plementation Specification. OpenGeospatial Consortium. Retrieved from http://portal.opengeospatial.org/files/?artifact_id=3839&version=1. Neis, P., Zipf, A. 2007. Zur Kopplung von OpenSource , OpenLS und OpenStreetMaps in OpenRouteService.org. G. Cugola and A. Margara. 2011. Processing flows of information: From data stream to complex event processing. ACM Computing Survey. Mobasheri, A. 2013. Exploring the possibility of semi-automated quality evaluation of spatial datasets in Spatial Data Infrastructure. Journal of ICT Research and Applications, 7(1), pp. 1-14. Mobasheri, A., Bakillah, M., Zipf, A., Liang, S.H.L. 2013. QualEvs4Geo: A Peer-to-Peer system architecture for semiautomated quality evaluation of geo-data in SDI. Third IEEE International Conference on Innovative Computing Technology (INTECH’13), London, UK.
