Finding OERs with Social-Semantic search

Finding OERs with Social-Semantic search
of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  See discussions, stats, and author profiles for this publication at: Finding OERs with Social-Semantic search Conference Paper  · May 2011 DOI: 10.1109/EDUCON.2011.5773299 · Source: IEEE Xplore CITATIONS 6 READS 95 5 authors , including: Some of the authors of this publication are also working on these related projects: Smart Land View projectSmart water network View projectNelson PiedraUniversidad Técnica Particular de Loja 56   PUBLICATIONS   124   CITATIONS   SEE PROFILE Janneth ChicaizaUniversidad Técnica Particular de Loja 44   PUBLICATIONS   114   CITATIONS   SEE PROFILE J. LópezUniversidad Técnica Particular de Loja 39   PUBLICATIONS   104   CITATIONS   SEE PROFILE Edmundo TovarUniversidad Politécnica de Madrid 186   PUBLICATIONS   489   CITATIONS   SEE PROFILE All content following this page was uploaded by Nelson Piedra on 26 February 2014.The user has requested enhancement of the downloaded file.    Finding OERs with Social-Semantic Search  Nelson Piedra, Janneth Chicaiza, Jorge López Escuela de Ciencias de la Computación Universidad Técnica Particular de Loja, UTPL Loja, Ecuador [nopiedra|jachicaiza|jalopez] Edmundo Tovar Facultad de Informática Universidad Politécnica de Madrid, UPM Madrid, España Oscar Martinez Centro de Investigación Operativa Universidad Miguel Hernández, UMH Elche, España  Abstract   —Social and semantic web can be complementary approaches searching web resources. This cooperative approach lets enable a semantic search engine to find accurate results and annotate web resources. This work develops the components of a Social-Semantic search architecture proposed by the authors to find open educational resources (OER). By means of metadata enrichment and logic inference, OER consumers get more precise results from general search engines. The Search prototype has been applied to find OER related with computers and engineering in a domain provided by OpenCourseWare materials from two universities, MIT and UTPL. The semantic search answers reasonably well the queta collected.  Keywords-component; Open Educational Resources, OER,  Semantic Web technology, searching architecture, social Web, knowledge domains, social-semantic Web paradigm, social annotations, semantic metadata technology, information seekers, ontology, OCW I.   I  NTRODUCTION 1  One of the reasons for which Social Web or Web 2.0  became so popular is that it is focused on contents, relations and knowledge and not precisely on technology [1]. Web 2.0 technologies augment the Web power. On the other hand, the Semantic Web vision [1] has inspired a large community of researchers and practitioners, who have achieved several early successes in the past six years. Semantic technologies enable structure and semantically enriching the content - to define and use common vocabulary, to generate a new knowledge or to solve word-meaning  problems-. These and other capabilities made it possible to improve or automate certain tasks that a human agent would not be able to perform. The semantic web has an inner capability for processing a large amount of information. The synergy between social and semantic web instances is analyzed by Gruber [3]: “The social web is an ecosystem of  participation, where value is created by the aggregation of many individual user contributions. The semantic web is an We would like to thank the Comunidad de Madrid for the support of EMadrid  Network of Excellence S2009 TIC-1650. The authors would also like to thank the comments by the eMadrid partners to the paper. This work has been also  partially funded by scholarship provided by the “Secretaría Nacional de Ciencia y Tecnología” of Ecuador (SENACYT) and with support of the  project AL11-P(I+D)-17 (UPM). ecosystem of data, where value is created by the integration of structured data from many sources”. Open Educational Resources 2  (OER) is the simple idea that the world’s knowledge is a public good and that technology in general and the Web in particular provide an extraordinary opportunity for everyone to share, use, and reuse knowledge. This paper is essentially an applied version to OER search  based on the Social-Semantic Search Architecture of [4]. The  purpose of this article was the semantic search of Requirements Engineering Artifacts. Now, we aim to extend the search domain to the OER. The section II describes how the semantic and social web could contribute to the search of OER. The section III shows some developed elements (architecture, ontology, user interface), and tasks (loading, quality assurance) used for the implementation of the OER search engine. The section IV describes an experimental evaluation and results discussion. Finally, conclusions are presented in section V. II.   S OCIAL -S EMANTIC T ECHNOLOGIES DRIVEN TO OER    S EARCHER Social and Semantic Web are two approaches complementary and each must draw from the other’s strengths [5]. In this regard, the ontology metadata provides the benefit of enabling a semantic search engine to find accurate results and to apply reasoning procedures on the metadata [6]. Respect to the social dimension, Wu et al. in [7] states that social annotations remove the high barrier to entry because web users can annotate web resources easily and freely; it directly reflects the dynamics of the vocabularies of the users and thus evolves with the users. This cooperative approach is called, social-semantic web. To Torniai et al. [8] it will let ‘creating, managing and sharing information through combining the technologies and approaches from Web 2.0 and the Semantic Web’. Merging the  best of both approaches can play a crucial role in the OER search [9][10]: semantic enrichment of tags or content created  by users through social tools [11], [12] social annotations for recommend social systems and folksonomies to populate an ontology. 2  OER is a special term used for learning materials which are open and freely licensed. OERs are most easily identified by their use of certain license types, especially Creative Commons license types. 978-1-61284-643-9/11/$26.00 ©2011 IEEE2011 IEEE Global Engineering Education Conference (EDUCON) – "Learning Environments and Ecosystems in Engineering Education"April 4 - 6, 2010, Amman, JordanPage 1195     A.   Vertical Search engines for OER The general search engines which are based on horizontal search, do not know the context, “because content is not capable to give systems structured and explicit information about itself in order to be better organized or indexed” [3]. Searching of Educational Resources, the development and usage of vertical search engines (VSEs) could be the option to find more accurate results, thus: specialized search (to find out and retrieve only educational content) filtering of content (to focus the target audience and its preferences), and, the possible to choose the most key educational resources providers.  B.    Related works In WikiEducator  , a list of specialized search engines and services are shown. Tools have been set up to make finding OER easier. Each search engine has different strengths and may not provide the accuracy and speed of standard web searches”. Between the most known or trustable are:  Learning  3 , OERs Commons 4 , OCWFinder  5   OER  Recommender  6  and  Discovered  7  prototype. From all we have mentioned, the last three ones, incorporate social and/or semantic technologies, as follows: •   OCWFinder   and OER Recommender   are part of  Folksemantic 8  project. Through this tools, students, educators and learners can receive personalized recommendations. Users can comment on and share resources with others as well as receive personalized recommendations based on their interests. •    Discovered   is a search prototype developed in the ccLearn division (of Creative Commons) [14]. This tool takes advantage from structured data (metadata) from educational resources to get an “enhanced search” of OERs, and although it is on its Beta Version, one of the best consuming metadata described in RDFa. •   Courseware Watchdog  , a tool based on ontologies and free code usable to manage courses. Tane in [[15] “show how ontologies help finding and organizing distributed courseware resources by offering a common framework for the retrieval and organization of courseware material”. In this work, by means of metadata enrichment and logic inference, OER consumers will get more precise results from general search engines. And at the moment of determining the relationship between the open resources, social annotations, RDF graphs and expert's recommendations, the system itself will be in charge of recommending action paths for information seekers. 3 4 5 6 7 8  Folksemantic is a project that “uses social networking to encourage user interaction and user generated content” III.   S YSTEM DESCRIPTION  The components of the system described in the following sections are: the searching architecture, the domain of different data sources, the definition of the ontology used, generic tasks that are related to ensuring semantic metadata quality and which should be supported by semantic web infrastructure design, loading metadata of the Open Academic Content Providers selected, and the search interface. Finally, the technologies used to implement the search prototype are shown.  A.    Description of the Searching Architecture When people search in the Web, specific situations are frequently found in [16] and [17]. The architecture proposed in [4] describes 5 principles and 10 criteria that support the search engine design. The Architecture includes 3 areas of the knowledge management: acquisition  (modules: metadata reader and annotation finder), encoding and storage  (modules: metadata processor, RDF triplets generated and stored), and knowledge application  (modules: query engine using SPARQL, Inference engine and feedback user system).  B.   The Context In the case OER search the data relevant for building the application is spread in several different data sources such as OCW sites, Social OER repositories, and HTML pages. Information about open educational content includes: textbooks, papers, presentations, source material, video/audio tutorials or demonstrations on specific topics, entire online courses (e.g. OCW), teacher and student guides lectures that are created and shared online is maintained in heterogeneous sites/repositories. For purposes of the publication scope, information about OER is maintained and stored in RDF Store with metadata obtained from two institutional open repositories, MIT ( and UTPL ( Beside its heterogeneous nature, another important feature of the OER domain is that the content offered on the web sites continuously undergoes changes. The RDF store change to reflect additions of new OER. It is planned to consume new OER through RSS. New contents are queried through RSS feeds and added to the RDF store. Therefore, the semantic data that underlies the application has to be often updated. OER are freely available materials — including educational content in a variety of forms and tools for learning and managing the work of education — provided or designed with the intent of being shared at little or no cost. In addition to the above, issues how community 9 , collaboration, sharing, and availability are key features of OER. C.   Ontology used We decided to use and extend an existing ontology from a  previous work [18] for representing the semantic metadata of 9  Groups such as the Open CourseWare Consortium ( encourage collaboration and sharing as well as discussion about OER/OCW 978-1-61284-643-9/11/$26.00 ©2011 IEEE2011 IEEE Global Engineering Education Conference (EDUCON) – "Learning Environments and Ecosystems in Engineering Education"April 4 - 6, 2010, Amman, JordanPage 1196    OER. We use the OER-CC ontology 10  reference ontology, which has been designed to support the OER search based on Semantic-Social Web technologies. We extended this ontology  by adding some domain specific concepts, such as licenses, etc. This ontology considering the use of ccREL standard [19] which provides information related to Licenses Properties, e.g.,  permissions, prohibitions, requirements and general metadata of a resource based on Dublin Core and IEEE-LOM. TABLE I. G LOSSARY E XCERPT OF T ERMS FOR OER    G ENERAL I  NFORMATION   Name Synonyms Description Type Open Educational Resource OER: Open Learning content Open Teaching and Learning Material OER is a digitalized material offered freely and openly for educators, students and self-learners to use and reuse for teaching, learning and research Concept entry Id, Identifier, OER URL The value of the OER identifier, i.e. the URL that is used to access an OER. Attributedescription - A textual description of the content of OER Attributekeyword - A keyword or phrase describing the topic of this OER Attributetitle - Name given to this OER Attributelocation URL An alternative URL that is used to access an OER Attributehas Format file format Property that links a OER to a file format Relation Format Format The concept that contains the  physical file format of the OER Concept file Format - A kind of OER format Attributemime type - A mime type is used to identify the type of information that a file contains Attributehas Language OER Language The primary human language or languages used within OER to communicate to the intended user. Relation Language Lang: X A Set of human language Concept code Language Code Code for the representation of names of languages Attributename Language  Name  Name of the language used to describe an OER Attribute  Through an iterative and collaborative process, the ontology adaptation was performed. This process was guided  by proposed activities in [4] using Methontology [20] and the use of graphic tool COE Cmaptools 11 . An excerpt of Glossary is presented in TABLE I. and an example of axiom is showed in TABLE II. TABLE II. A  N EXAMPLE OF AXIOM FOR OERCC  ONTOLOGY   Axiom name LearningResurceType-fileFormat Description Nowhere OER of simulation type may be associated to a file format based on text. 10 11 Expression forall(?X)([OpenEducationalResource](?X) and [hasEducationalInformation(?X, ?Y) and [hasLearningResourceType](?X, "Simulation") and [format](?X, ?Z) and [fileFormat](?Z, ?W)-> [different(?W, "text")] Concepts OpenEducationalResource Referred attributes Fileformat Ad hoc binary relations hasEducationalInformation, hasLearningResourceType, format Variables ?X, ?Y, ?Z, ?W  D.   Tasks for Ensuring Metadata Quality of the Semantic Metadata Based on the characteristics of a semantic search engine [21], we identify generic tasks that are related to ensuring semantic metadata quality and which should be supported by semantic web infrastructure design in [4]. In our application we  provide support for all these generic quality related tasks: •   Development automatic methods to extract the required metadata from heterogeneous data sources. It is important to employ automatic methods so that the  process can be easily repeated periodically in order to keep the knowledge updated. Another important characteristic of the extraction mechanism is that it should be adaptable to the content of the sources that are explored. •   Ensure that the derived metadata is free of common errors: different identifiers that refer to the same entity (institutions, knowledge areas, authors, etc), instances whose meaning is not clear, entity which needs to be disambiguated 12 . Errors can be caused by data entry mistakes, by information extraction tools, or by the inconsistency and duplication entries of diverse data sources. •   Update the semantic metadata as new information  becomes available so that the knowledge base is updated in an appropriate fashion. In the following section we overview a set of semantic web components that rely on data acquisition and integration and describe how they approach to our application.  E.    Loading Metadata to RDF Store The process starts with the selection of Open Academic Content Providers. Two providers were selected: OpenCourseWare from MIT, and the university where the work is being carried out, OCW-UTPL 12  Two major approaches to avoid these errors are first, attempt to tackle these errors before the data has been extracted (e.g., using lexicons); The second way to approach this problem is to clean the semantic data after it has been extracted. This approach requires mechanisms to correctly diagnose the  problem at hand and then algorithms to correct each individual problem. 978-1-61284-643-9/11/$26.00 ©2011 IEEE2011 IEEE Global Engineering Education Conference (EDUCON) – "Learning Environments and Ecosystems in Engineering Education"April 4 - 6, 2010, Amman, JordanPage 1197    Figure 1. Metadata extraction proc Different criteria can be considered foinformation providers’ selection, like for exaauthor or creator, comparability with related information, purpose and audience. Once the been identified, Figure 1. shows the developcan observe, different activities are into the to integrate traditional databases: Extract, Tr (ETL). Figure 2. Interface of the search engine: Tags, The extraction sub process is made autocomponent called query engine which bdifferent SPARQL queries on RSS files of obtain: •   General information about OER:  publisher, language, rights, date, (calculated). •   Subjects or area of knowledge to educational resource. •   Resource creators’ name For each type of query it generates a typ metadata , which is pre-processed and debugorder to remove or replace special character    ess r the content and mple: authority of source, stability of data sources have d process. As we eneral stages used ansform and Load social ranking atically through a asically, executes CW providers to title, description, URL and URI hich belongs the e of file: resource ed step by step, in and in general to verify data quality. During the files –generated during the transformed by a written progconsiders the structure and vontology [18], to generate the NT format. TABLE III. OER Provider OCW-MIT Knowledge area: ElectricalEngineering and Computer Science Address Web Feed: feed:// OcwWeb/rss/new/mit-newcourses-6 UTPL-OCW Knowledge area: Systems and Computer Engineering Address Web Feed: feed:// Finally, through the load offered by the database managthis moment, in a SQL Endexecuted by whoever knows students, teachers, or self-leasearch client to find resources TABLE III. Error! found. summarizes the repositcarry out this process. Figure 3. Tier Layers of t  F.   Search Interface Search interface is showedeach found resource is presentsecond stage, each of the three revious stage-, are read and am in JAVA and JENA, which cabulary defined by OER-CC correspondent RDF triplets in OER    P ROVIDERS   Obtained Metadata Resource URL, title, description, Publisher, language, rights, date, creator, relation and subject Kind of resources: OCW courses Triplets number:  1380 Resource URL, title, description, Publisher, language, Rights, date, creator and subject Kind of resources: OCW courses and Youtube videos  Triplets number: 1690 manager and data publication r, the RDF file is loaded. From oint, different queries can be the query language SPARQL; ners who could use the Web f their own interest. Reference source not ries and resources necessary to he Search Engine Prototype in Figure 2. and Figure 4. For ed the tags with the ones it has 978-1-61284-643-9/11/$26.00 ©2011 IEEE2011 IEEE Global Engineering Education Conference (EDUCON) – "Learning Environments and Ecosystems in Engineering Education"April 4 - 6, 2010, Amman, JordanPage 1198
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!