Internet & Technology

A multimedia broker for accessible learning objects transcoding

Description
A multimedia broker for accessible learning objects transcoding
Published
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
Share
Transcript
  See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/224645262 A Multimedia Broker for Accessible LearningObjects Transcoding Conference Paper  · January 2006 DOI: 10.1109/ICTTA.2006.1684447 · Source: IEEE Xplore CITATIONS 4 READS 14 4 authors , including:Silvia MirriUniversity of Bologna 66   PUBLICATIONS   252   CITATIONS   SEE PROFILE Marco RoccettiUniversity of Bologna 221   PUBLICATIONS   1,733   CITATIONS   SEE PROFILE Paola SalomoniUniversity of Bologna 112   PUBLICATIONS   456   CITATIONS   SEE PROFILE All content following this page was uploaded by Silvia Mirri on 08 October 2015. The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the srcinal documentand are linked to publications on ResearchGate, letting you access and read them immediately.    MultimediaBroker for Accessible Learning Objects Transcoding Silvia Mirri, Daniele Pantieri, Marco Roccetti, Paola SalomoniDepartment ofComputer Science   University of Bologna (Italy) Mura Anteo Zamboni, 7- +390512094516, tmirri, pantieri, roccetti, Abstract Thewide diffusion of e-learningtechnologiesrepresents a great opportunity forunderserved segmentof the population. This is particularly true for people with disabilities, whomay have difficulties attending traditionalon-site learning programs or accessing to conventional (printed) learning materials. The creation and provision of accessible e-learning contents are crucialaspects in ensuring that e-learning will become an inclusive technology. This paper presents LOT (Learning Objects Transcoder) a system designed and developed to automatically produce diferent accessiblelportable versions of audiolvideo learning materials. The system offers a broking service to transcode digital video lectures included in a LO thatfullfits a specific student profile, focusing specifically on hislheraccessibility requirements. 1. Introduction Offering an increasing access to more learners is usuallyconsidered one of the main benefits provided by e-learning but frequently on-line educational and training services are based on anytime and anywhere technologies that are not for everyone. Commonly e- learning materials are designed to be used with a specific hardware device, with a particular software technology and an exact configuration so that they are less available to people who have limited access capabilities and specifically to students with disabilities. As a result, today e-learning is one of the more inaccessible Web- based technology and students with disabilities are frequently ruled out from virtual classrooms. This is particularly crucial because learners with disabilities can add to usual benefits of e-learning some other positive effects derived from their specific needs [1, 2]. We can consider, for example, students with visual impairments who have difficultiesin reading traditional printed materials or people with motion disabilities who may have problems in attending on-site programs. Some otherconsiderations force e-learning platforms and contents to become accessible. First, different countries have encouraged or compelled accessibility by law, like, for example, US [3] and Italy [4]. Second, improving accessibility of learning resources can also resultin an - 4012 7 Bologna (Italy) +390512094510 salomoni},cs. unibo. it improved e-learning experience for everyone. Moreover, making e-learning accessible ensures thatdidactical materials can beused by all studentsregardless of environmental or technological constraints, sustaining the accommodation of individual learning styles and preferences [5]. Differentstandards and guidelines [6, 7, 8] have been proposed to effectively support accessible e-learning (and accessible Web pages [9] but the provision of fully accessible and portable contents is usuallyconsidered as a time-consuming anddismaying task.In particular, thethree main IMS specifications on accessibility have driven our work: * IMS Guidelines for Developing Accessible Learning Applications [6], that canbe used to support the production of accessible learning materials. a IMS Accessibility forLearner Information Package(ACCLIP) [7], to describe personal profile of preferences (visual,aural, device) fortailored presentation of learning content (e.g. preferred/requiredinput/outputdevices or preferred content alternatives). a IMS AccessForAll Meta-data [8], that describes accessible learning content specifying, for example, what kindof content is going to be presented and if there is an equivalent or alternative form of the content. The AccessForAllMeta-data specification is intended to make it possible to identify resources that match a user's stated preference or need, by working on the LO manifest that is a standard part of SCORM [10]. These preferences or needs would be declared using the IMS Learner Information Package Accessibility for LIP specification.All the abovementioned issues are particularly crucial due to the wide use of rich media resources done in creating learning materials. Inthis context we present LOT (Learning Objects Transcoder), a system designed and developed to use transcoding strategies in automatically producing accessible LOs. The term  Transcoding signifiesthe process of converting a media file or object from one format to another. LOT works on a repository of audio and video lecturesthat synchronously combine different continuous media andproduces a personalized and re-purposed learning object on thebasis of the LIP profile provided by the user  or, 0-7803-9521-2/06/ 20.00 §2006 IEEE. 641  more generally, by the client system). The srcinal content is typically flexible in terms of how it is reproduced and controlled. It is also created with all equivalents needed to ensure multimodality. LOT uses this raw material to automatically create the adapted version needed by the student to proficiently use the learning materials. LOT also creates the SCORM [10] format by including also all accessibilityrelated meta-data. Hence, LOT supports learning contents accessibility automatically producing different accessible/portableversions of audio/video learning materials. All contents are implemented as SCORM LOs, ensuring accessibility issues as well asportability through delivery systems. The remainder of the paper is organized as follows. Section 2 describes main design issues and illustrates how they have influenced the design of system architecture. Section 3 presentsacase study, giving emphasis to profiling and transcoding activities performed byLOT. Finally Section 4 concludes the paper. 2. Design issues Our system is basically a broking service that uses a studentprofile to appropriatelytranscode digital video lectures included in a SCORM [10] LO andcoded in SMIL [11]. Original resources (including all audio and video files are transformed by LOT in the form that full fits specificstudent needs,focusing on his/heraccessibility requirements. In order to profile students with disabilitiesin e-learning provisioning we used IMS ACCLIP Specification [7] that provides a means of describing how learners interact with an e-leaming environment. The ACCLIP Specification only deals with accessibility preferences and is apart of a more complete IMS Learner Information Package (LIP) Specification. It defines the elements required in order to representaccessibility preferences, grouped into four sections  i Display information,  ii Control information,  iii Content information and (iv) Accommodations for which a learner is eligible. Themain idea that has driven our work is to manage SCORM package in order to  i un-package requested contents,  ii to transcode them in a suitable way,  iii to add (or simply modify) related meta-data and then (iv) to re-package trancoded contents obtaining a new LO. Hence aclientthatcontacts our system sends a student profile and a request of a LO and obtains back theaccessible LO corresponding to the request, optimized for the requiredprofile and appropriatelyencapsulated. To perform this activity our system acts into three steps: * Broking, the system performs a broking role between theavailable resources and the user profile [12]to find the better formof the whole video lecture for aspecific user. * Transcoding, the system orchestratesa set of specific transcoding Web Services to obtain the requiredversion of the entire video lecture. * Packaging, the system encapsulates theaccessible video lecture by appropriately adding meta-data required by the mentionedabove e-learning standards [7, 8, 10]. The general architecture of our system may be thought as constructed out of the following software components, depicted in Figure 1: * The Media Broker (MB) manages users' access to thesystem. It provides means for user registration and authentication via user name and password during the first access, thus allowing transcoding facilities for authorized users. But mainly, MB is responsible of agreeing on theprofile definition at the start up ofeach connection and then specifying and controlling the transcoding activity. Choices about what and how to transcode are taken on thebasis of profile and contents descriptions. * The Profile Manager (PM) controlsthe IMS ACCLIP Profiles Data Base, maintaining user accessibility preferences and needs. Once auser is authenticated to the Media Broker, it gets accessibility profiles by querying the PM. * ThePackage Manager (PaM) handles SCORM packages, un-packaging requested LOs, before the transcodingphase, in order to transform separately each media contents. Once the transcoding phase is completed, the PaM re-composes the video lecture by combining single media. Then, the PaM adds appropriate meta-data (including IMS AccessForAll Meta-data [8]) to the new video lecture to obtain a SCORM [10] compliant LO. * The Transcoding Unit (TU), that essentially transcodes the video lecture. In order to improve system performances TU has been implemented as a two-level Web Services sub-component. In particular, the first level is composed by a specific Web Service that is embedded within TU,whichmanages the SMIL specification and transcodes it. The resulting mark-up (SMIL or XHTML) requires a set of sub-transcoding in order to complete the media transformations. These transformations are performed by engaging satellite Web Services which constitute the second level Web Services. These single transcoding units perform simple transformations on single media that could be computationallyheavy, i.e. transforming a video from a specific size to another. They could also be engaged from  well known Web Services available on the net to perform a specific task, i.e. a translation service (our system is currently using www.webservicex.net to translate text [13]). So, this second level consists of satellite Web Services, which may be local or external ones. External Web Services are deployed over theInternet 0-7803-9521-2/06/ 20.00 §2006 IEEE. 642    IMS ACCLIP Profiles  Jr Figure 1. LOT Architecture and can be contacted only when TU identifiesthe need for a particular transformation skill which is not offered by local Web Services. On the other hand, local Web Services are locally deployed (on the same LAN of the TU first level) and they perform specific transcodings. Figure 1 illustrates the whole flow and depictsthe interaction among the described components, as described in the following. As soon as the user issues a request for a particular Learning Object, the client application contacts LOT by authenticating itself to the MB. In correspondence to the first access of the user to the LOT, the user specifies accessibility preferences and related needs, and then the system identifies the profile by a unique ID. All such information is stored within the PM. Then, during every following access to the LOT, the user simply specifies just this unique ID. Thus, as soon as a requestarrives at the MB, it interrogates the PM and retrieves the user'sprofile. Then, the MB defines a transcoding strategy, by matching user's profile and the specific video lecture resources, and passes such information to the TU, calling the Package Manager in order to un-package the requested LO. The TU receives thesrcinal media sources and computes a transcoded media version, according with the requirementsspecified by the MB and theaccessibilityconstrains described in [7]. The TU is a sub-componentwhose task is dual. First, it schedules the transcoding activities on thebasis of the transcoding plan, needed for an optimal delivery of the presentation to the user and identified by the MB. Second, it embeds some services designed to locally accomplish specific transcoding processes. Inparticular, the TU is composed by two differentlevels of Web Services:the first one (that is the TranscodingUnit Web Service) performs the whole video lecture resource transcoding and identifies the needsof single media transcoding, engaging the appropriate Web Service. These second level Web Services are devoted to the transformation of a specific media type. Each Web Service performs adifferenttranslation and re-coding activity, ranging from simple re-sizing of fonts in order to display them on the screen, to complex mechanisms that transform video flow to exclusively audio flow.Finally, TU passes all the transcoded resources to the PackageManager that provides appropriatestandard meta-data to the resources and then re-packages these ones in a new SCORM package. Finally, it sends the adapted LO back to the client. System performances are improved using a two-level caching system, depicted in figure 1. The first level Transcoding Unit is supplied with first level cache that maintains recently managed files, both LO and SMIL/XHTML structures.   second level cache is provided to manage recently used files that are transcoded by eachsecond level TU. In these caches are storedthe single media files produced by a specific second level TU. The LOT takes advantage of caching systems by transcoding resources datajust once and reusing it by re-delivering the file more times to users with similar profiles [ 14]. 0-7803-9521-2/06/ 20.00 §2006 IEEE. II 643  3. Experimental Scenarios In this section we illustrate a benchmark case study in which aspecific LO (containing a video lecture) is transcoded by LOT on thebasis of two different profiles. The srcinal LO, a synchronizedvideo lecture, is described in section 3.1 while Section 3.2 illustrates the two profiles used for thecase study. Finally in Section 3.3 we presentthe two LOs corresponding to the considered profiles and we briefly describe the process engaged by LOT to produce them. 3.1. Original LO description We consider an srcinal resource represented by a digital multimedia lecture, created by using a SMIL-based synchronizedmultimedia content. SMIL [ 1] is a set of XML moduleswhich are used to describe the temporal, positional, and interactive behavior of a multimedia presentation. The SMIL mark-up language can be used to manage timing, layout and animation in compositions ofsynchronous multimedia.This srcinal digital multimedia lecture is produced with the SMIL version 2.0 and it is compliant to SMIL accessibilityfeatures [15]. Moreover, the lecture is distributed and packaged in compliance with SCORM standard(version 1.3) [10]. The SCORM package contains meta-information structured like IMS [8] and SCORM meta-data. Figure 2. The synchronized multimedia lecture The digital multimedia lecture is composed of three synchronized flows of information [16]: 1.   video track, showing the lecturer. 2. An audio track, playing the lecturer's talk. 3.   sequence of slides, consisting of a sequence of images with text, charts or other graphical elements that help the student to follow the lecture. To ensure portability and accessibility, two other informationflows are also addedand maintained synchronized with other ones: 1.   caption sequence; captioning is essentialfor deaf and hard ofhearing students, as well as foreign students. The use of caption is necessary also to provide support for students thatare using devices without audio output capabilities; 2. An additionaldescription of contents for each slide being presented;these additional information can be exploited as an alternative information to the mediacomposing the lecture.   screenshot of a lecture developed as described (in Italian language), is depicted in Figure 2. The lecture is displayed with captioning set on. Each portion of therich media content (corresponding to one slide) is described by a SMIL code fragment, as described by Figure 3. All audio and text files are in Italian. Figure 3. The synchronized video lecture Figure 3. The synchronized video lecture (corresponding to a slide) 3.2. User profiling In the following, we describe two profiles defined in order to test the architecture of LOT mentioned in the previous section. In the first one (A), we consider a scenario in which the user is a blind person, that gains access to the multimedia lecture from home, using a PC equipped with ascreen reader Jaws 5.0 [17]  i.e., the assistive technology that enables unsighted people to use a computer)and a Braille display. In the second case (B), we assumed a scenario in which a typical useraccesses to LOT systemwith a fully equipped PC  i.e., with full support for SMIL and high quality multimedia flows). The preferences of   and B and the capabilities of the devices they are using, are collected in two profiles in compliance with ACCLIP standard. 0-7803-9521-2/06/ 20.00 )2006 IEEE. <par> <video src= video/video.rm region= region_video /> <seq><par> <imgregion= region_slide src= img/l.gif dur= 60s height= 361 width= 480 alt= Fondamenti di ElaborazioneMultimediale longdesc= img/l.htm /> <audio region= region audio src= audio/l.mp3 /> <textstream src= caption/l.rt region= region_subtitle width= 640 system-captions= on /> </par></seq></par> 644
Search
Tags
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