No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Learning Driver Preferences of POIs Using a Semantic Web Knowledge System
Abstract
In this paper, we present the architecture and implementation of a Semantic Web Knowledge System that is employed to learn driver preferences for Points of Interest (POIs) using a content based approach. Initially, implicit & explicit feedback is collected from drivers about the places that they like. Data about these places is then retrieved from web sources and a POI preference model is built using machine learning algorithms. At a future time, when the driver searches for places that he/she might want to go to, his/her learnt model is used to personalize the result. The data that is used to learn preferences is represented as Linked Data with the help of a POI ontology, and retrieved from multiple POI search services by ‘lifting' it into RDF. This structured data is then combined with driver context and fed into a machine learning algorithm that produces a statistical model of the driver's preferences. This data and model is hosted in the cloud and is accessible via intelligent services and an access control mechanism to a client device like an in-vehicle navigation system. We describe the design and implementation of such a system that is currently in-use to study how a driver's preferences can be modeled by the vehicle platform and utilized for POI recommendations.
... which is a restaurant recommender, the Sightsplanner (tallinn.sightsplanner.com) for Tallinn related recommendations and the Supe system (Parundekar & Oguchi, 2012), which personalises the GPS devices of the car drivers. ...
... Thus, this data could be reliable to extract indications about the preferences of the users. Moreover, as presented in Section 2, the Supe system exclusively exploits the visited POIs of the drivers in order to recommend POIs in the future (Parundekar & Oguchi, 2012), which amplifies the idea of exploiting the personal data in LBSN networks. ...
According to recent studies and forecasts in the field of computer science, a significant portion of on-going research is related to the data that is generated and made available on the Web; this data is also known as ‘big data’ due to the volume of the storage space needed to retain it. This study aims to shed some light on the relation between the personal data that could be extracted from the Social Web and the various applications in the tourism domain that could profit from this personal data. The motivation is the design of an ecosystem that harnesses the power of shared personal data towards producing personalised services, offers and recommendations for users. One of the basic prerequisites, and the purpose of this investigation, is the creation of a user model specifically for the tourism sector, which aggregates the relevant tourism user properties.
The space inside a vehicle, which people can enjoy while travelling, is becoming more intelligent due to the technological developments within the automotive industry. This has influenced the increase in new services that are intended to fulfil the needs of vehicle users. This article presents a survey on the provisioning of ubiquitous intelligent services for vehicular users. It also identifies clusters of research interest on this subject. To this end, an evaluation of results from six scientific research databases supports these goals. This process initially identified 37 328 publications; after filtering and clustering, the scope of analysis became 39 articles. The main contributions points to (i) the existence of five active research clusters, as a result of the inclusion of a visual clustering step in the systematic mapping protocol; (ii) an indication of a string of studies starting in the late ’90s from traffic issues, intelligent transport systems and profiling, to smart cities, the Internet of Things, big data, fog computing and internet of vehicles in recent years; and (iii) that the main challenges associated with the implementation of ubiquitous intelligent services for vehicular users are related to data security, infrastructure, connectivity and high mobility.
In this paper, we present an approach for automatically recommending categories for spatiotextual entities, based on already existing annotated entities. Our goal is to facilitate the annotation process in crowdsourcing map initiatives such as OpenStreetMap, so that more accurate annotations are produced for the newly created spatial entities, while at the same time increasing the reuse of already existing tags. We define and construct a set of training features to represent the attributes of the spatiotextual entities and to capture their relation with the categories they are annotated with. These features include spatial, textual and semantic properties of the entities. We evaluate four different approaches, namely SVM, kNN, clustering+SVM and clustering+kNN, on several combinations of the defined training features and we examine which configurations of the algorithms achieve the best results. The presented work is deployed in OSMRec, a plugin for the JOSM tool that is commonly used for editing content in OpenStreetMap.
This document is obsolete. The definitive document is Standard ECMA-404 The JSON Data Interchange Syntax.
JavaScript Object Notation (JSON) is a lightweight, text-based, language-independent data interchange format. It was derived from the ECMAScript Programming Language Standard. JSON defines a small set of formatting rules for the portable representation of structured data.
The structure of the Semantic Web gives users the power to share and collaboratively generate decentralized linked data. In many cases, though, collaboration requires some form of authentication and authorization to ensure the security and integrity of the data being generated. Traditional authorization systems that rely on centralized databases are insufficient in this scenario, since they rely on the exis-tence of a central authority that is not available on the Semantic Web. In this paper, we present a scalable system that allows for decentralized user authentication and authorization. The system described supports per-document access control via an RDF metadata file containing an access control list (ACL). A simple interface allows authorized users to view and edit the RDF ACL directly in a Web browser. The system allows users to efficiently manage read and write access to linked data without a centralized authority, enabling a collaborative authoring environment suited to the Semantic Web.
The Semantic Web can be viewed as one large "universal" RDF graph distributed across many Web pages. This is an impractical for many reasons, so we usually work with a decomposition into RDF docu-ments, each of which corresponds to an individual Web page. While this is natural and appropriate for most tasks, it is still too coarse for some. For example, many RDF documents may redundantly contain the same data and some documents comprise large amounts of weakly-related or unrelated data. Decomposing a document into its RDF triples is usually too fine a decomposition, information may be lost if the graph contains blank nodes. We define an intermediate decomposition of an RDF graph G into a set of RDF "molecules", each of which is a connected sub-graph of the original. The decomposition is "lossless" in that the molecules can be recombined to yield G even if their blank nodes IDs are "standardized apart". RDF molecules provide a useful granularity for tracking the provenance of or evidence for information found in an RDF graph. Doing so at the document level (e.g., find other documents with identical graphs) may find too few matches. Working at the triple level will just fail for any triples containing blank nodes. RDF molecules are the finest granularity at which we can do this without loss of information. We define the RDF molecule concept in more detail, describe an algorithm to decompose an RDF graph into its molecules, and show how these can be used to find evidence to support the original graph. The decomposition algorithm and the provenance application have both been prototyped in a simple Web-based demonstration.
The increasing availability of large RDF datasets offers an exciting opportunity to use such data to build predictive models using machine learning algorithms. However, the massive size and distributed nature of RDF data calls for approaches to learning from RDF data in a setting where the data can be accessed only through a query interface, e.g., the SPARQL endpoint of the RDF store. In applications where the data are subject to frequent updates, there is a need for algorithms that allow the predictive model to be incrementally updated in response to changes in the data. Furthermore, in some applications, the attributes that are relevant for specific prediction tasks are not known a priori and hence need to be discovered by the algorithm. We present an approach to learning Relational Bayesian Classifiers (RBCs) from RDF data that addresses such scenarios. Specifically, we show how to build RBCs from RDF data using statistical queries through the SPARQL endpoint of the RDF store. We compare the communication complexity of our algorithm with one that requires direct centralized access to the data and hence has to retrieve the entire RDF dataset from the remote location for pro- cessing. We establish the conditions under which the RBC models can be incrementally updated in response to addition or deletion of RDF data. We show how our approach can be extended to the setting where the attributes that are relevant for prediction are not known a priori, by selectively crawling the RDF data for attributes of interest. We pro- vide open source implementation and evaluate the proposed approach on several large RDF datasets.
The work on integrating sources and services in the Seman- tic Web assumes that the data is either already represented in RDF or OWL or is available through a Semantic Web Service. In practice, there is a tremendous amount of data on the Web that is not available through the Semantic Web. In this paper we present an approach to automatically discover and create new Semantic Web Services. The idea behind this approach is to start with a set of known sources and the corre- sponding semantic descriptions of those sources and then discover similar sources, extract the data from those sources, build semantic descriptions of those sources, and then turn them into Semantic Web Services. We implemented an end-to-end solution to this problem in a system called Deimos and evaluated the system across ve dierent domains. The re- sults demonstrate that the system can automatically discover, learn se- mantic descriptions, and build Semantic Web Services with only example sources and their descriptions as input.
Despite the increased awareness that exploiting the large amount of semantic data requires statistics-based inference capabilities, only little work can be found on this direction in the Semantic Web research. On semantic data, supervised approaches, particularly kernelbased Support Vector Machines (SVM), are promising. However, obtaining the right features to be used in kernels is an open problem because the amount of features that can be extracted from the complex structure of semantic data might be very large. Further, combining several kernels can help to deal with efficiency and data sparsity but creates the additional challenge of identifying and joining different subsets of features or kernels, respectively. In this work, we solve these two problems by employing the strategy of dynamic feature construction to compute a hypothesis, representing the relevant features for a set of examples. Then, a composite kernel is obtained from a set of clause kernels derived from components of the hypothesis. The learning of the hypothesis and kernel (s) is performed in an interleaving fashion. Based on experiments on real-world datasets, we show that the resulting relational kernel machine improves the SVM baseline.
The Web is gradually evolving as provider of services along with its text and image processing functions. Web services markup is proposed in the Defense advance research project agency's agent markup language (DAML) family of semantic Webmarkup languages. The markup provide an agent-independant declarative API to capture the data and metadata associated with a service. Sharing, reuse, composition, mapping and succint local Web service markup is facilitated by the exploitation of ontologies by markup. A wide variety of agent technologies for automated Web services discovery, execution, composition and interoperation is enabled by this markup.
This paper describes the theoretical background and the implementation of dbrec, a music recommendation system built on top
of DBpedia, offering recommendations for more than 39,000 bands and solo artists. We discuss the various challenges and lessons
learnt while building it, providing relevant insights for people developing applications consuming Linked Data. Furthermore,
we provide a user-centric evaluation of the system, notably by comparing it to last.fm.
We discuss algorithms for learning and revising user profiles that can determine which World Wide Web sites on a given topic would be interesting to a user. We describe the use of a naive Bayesian classifier for this task, and demonstrate that it can incrementally learn profiles from user feedback on the interestingness of Web sites. Furthermore, the Bayesian classifier may easily be extended to revise user provided profiles. In an experimental evaluation we compare the Bayesian classifier to computationally more intensive alternatives, and show that it performs at least as well as these approaches throughout a range of different domains. In addition, we empirically analyze the effects of providing the classifier with background knowledge in form of user defined profiles and examine the use of lexical knowledge for feature selection. We find that both approaches can substantially increase the prediction accuracy.
This chapter discusses content-based recommendation systems, i.e., systems that recommend an item to a user based upon a description of the item and a profile of the user's interests. Content-based recommendation systems may be used in a variety of domains ranging from recommending web pages, news articles, restaurants, television programs, and items for sale. Although the details of various systems differ, content-based recommendation systems share in common a means for describing the items that may be recommended, a means for creating a profile of the user that describes the types of items the user likes, and a means of comparing items to the user profile to determine what to recommend. The profile is often created and updated automatically in response to feedback on the desirability of items that have been presented to the user.
Web services are important for creating distributed applications on the Web. In fact, they're a key enabler for service-oriented architectures that focus on service reuse and interoperability. The World Wide Web Consortium (W3C) has recently finished work on two important standards for describing Web services the Web Services Description Language (WSDL) 2.0 and Semantic Annotations for WSDL and XML Schema (SAWSDL). Here, the authors discuss the latter, which is the first standard for adding semantics to Web service descriptions.
HP Labs developed the Jena toolkit to make it easier to develop applications that use the semantic Web information model and languages. Jena is a Java application programming interface that is available as an open-source download from www.hpl.hp.com/semweb/jena-top.html.
Design issues: Linked data
- T Berners-Lee
Berners-Lee, T.: Design issues: Linked data (2006),
http://www.w3.org/DesignIssues/LinkedData.html
Automatically Constructing Semantic Web Services from Online Sources
- J L Ambite
- S Darbha
- A Goel
- C A Knoblock
- K Lerman
- R Parundekar
- T Russ
- A Bernstein
- D R Karger
- T Heath
- L Feigenbaum
- D Maynard
- E Motta
Ambite, J.L., Darbha, S., Goel, A., Knoblock, C.A., Lerman, K., Parundekar, R.,
Russ, T.: Automatically Constructing Semantic Web Services from Online Sources.
In: Bernstein, A., Karger, D.R., Heath, T., Feigenbaum, L., Maynard, D., Motta,
E., Thirunarayan, K. (eds.) ISWC 2009. LNCS, vol. 5823, pp. 17-32. Springer,
Heidelberg (2009)
Relational Kernel Machines for Learning from Graph-Structured RDF Data
- V Bicer
- T Tran
- A Gossen
- G Antoniou
- M Grobelnik
- E Simperl
- B Parsia
- D Plexousakis
- P De Leenheer
Bicer, V., Tran, T., Gossen, A.: Relational Kernel Machines for Learning from
Graph-Structured RDF Data. In: Antoniou, G., Grobelnik, M., Simperl, E., Parsia,
B., Plexousakis, D., De Leenheer, P., Pan, J. (eds.) ESWC 2011, Part I. LNCS,
vol. 6643, pp. 47-62. Springer, Heidelberg (2011)
Tracking RDF graph provenance using RDF molecules
- L Ding
- T Finin
- Y Peng
- P Da Silva
- D Mcguinness
Ding, L., Finin, T., Peng, Y., Da Silva, P., McGuinness, D.: Tracking RDF graph
provenance using RDF molecules. In: Proc. of the 4th International Semantic Web
Conference, Poster (2005)
Using rdf metadata to enable access control on the social semantic web
- J Hollenbach
- J Presbrey
- T Berners-Lee
Hollenbach, J., Presbrey, J., Berners-Lee, T.: Using rdf metadata to enable access
control on the social semantic web. In: Workshop on Collaborative Construction,
Management and Linking of Structured Knowledge (2009)
Learning Relational Bayesian Classifiers from RDF Data
- H T Lin
- N Koul
- V Honavar
- L Aroyo
- C Welty
- H Alani
- J Taylor
- A Bernstein
- L Kagal
- N Noy
Lin, H.T., Koul, N., Honavar, V.: Learning Relational Bayesian Classifiers from
RDF Data. In: Aroyo, L., Welty, C., Alani, H., Taylor, J., Bernstein, A., Kagal, L.,
Noy, N., Blomqvist, E. (eds.) ISWC 2011, Part I. LNCS, vol. 7031, pp. 389-404.
Springer, Heidelberg (2011)
- A Passant
- P F Patel-Schneider
- Y Pan
- P Hitzler
- P Mika
- L Zhang
- J Z Pan
- I Horrocks
Passant, A.: dbrec -Music Recommendations Using DBpedia. In: Patel-Schneider,
P.F., Pan, Y., Hitzler, P., Mika, P., Zhang, L., Pan, J.Z., Horrocks, I., Glimm, B. (eds.)
ISWC 2010, Part II. LNCS, vol. 6497, pp. 209-224. Springer, Heidelberg (2010)