My Mini Project on Collaborative Filtering- A Concept Paper

Apr10

 

Juhaida Abdul Aziz, P53862

Abstract

Collaborative Filtering of Teaching in a Learning 3.0 Environment is indeed necessary especially for educators or teachers who considered technology and the Internet as the tools for their effective and meaningful teaching resources. As it is, teaching and learning collaboratively represent diverse life styles, cultures and religious in a spirit of mutual respond and open dialogue in personal, work and community in a discovery learning environment. In this concept paper, a quick glimpse based on previous researches, will touch on some prominent Web 2.0 applications in education, for instance, wikis, blogs and podcasts.  A quite significant revolutionary methods and ways in managing the multimodal online information and knowledge repositories, including the research information in teaching and learning of Web 2.0 technologies will be compared with the traditional Web 1.0 model.  The paper will look into Web 3.0 (the semantic web) and how it could be combined with 2.0 with it focal point to trigger the ultimate  production on Teaching in a Learning 3.0 Environment of ‘Read/ Write/ Collaborative Web’ which may drag the learners to unlimited boundaries in anytime. The ‘intelligent agents’ in software will allow the computers to filter out whatever unwanted and allow what the users want.    The collaborative filtering of teaching in this learning environment is yet to be considered to come into significant play to ensure that teaching and learning especially in Malaysian context and setting will be materialized accordingly and suit the National Philosophy of Malaysian Education.

Keywords: collaborative filtering, teaching, learning materials, 3.0 semantic web, intelligent agents

1.         Introduction

At a click of a mouse, World Wide Web enables users to search anything, anytime and anywhere without boundaries. Yet, too much information and materials to learn in the net, too much time to spend browsing to retrieve the right information to suit everyone needs and preferences, these are all significant factors where  Recommender systems are becoming more commonly used to help search the desired items accordingly. Recommender systems in e-learning can be differed in many ways depending on what kind of objects to be recommended; for instance, course to enrol, learning materials and etc. One of the proposed approaches used here is collaborative filtering, a system that can find users with similar interest and here we are focusing in teaching area in a learning 3.0 environment, in which new learning materials are to be recommended to teachers with a high degree of similarity. This system proposes a new e-learning framework using net services that has the ability to aggregate the recommended materials from other e-learning websites and predict more suitable materials to fulfill learners’ preferences. Along with that, the adaptive hypermedia system comes into play as this share the same goal as it aims to personalize the delivery of the materials according to the learners’ needs.

In this concept paper, section 2.0 touches on literature review of previous related studies followed by the evolution of web. Apart from this, discussions on the e-learning recommender system and adaptive hypermedia systems along with the collaborative filtering for teaching in current work will also be covered in this paper.

2.         Literature Review

2.1       Related Studies

Recent trend on e-learning recommender system shows that researches use several recommendation strategies namely, collaborative filtering, data mining techniques, content-based filtering, clustering, knowledge discovery, etc (Ghauth & Abdullah  2009). Table 1 illustrates the recommendation strategies proposed by current researchers on e-learning.

Table1. Recommendation strategies, input, and output of the current research

Researchers Recommendation
Strategies Input Output
Bandura (1997), Brusilovsky (2001) and Adomavicius (2005), Data mining techniques learner’s activities/ 

access history, learners

rating, item attributes

 related items/ 

documents,

related links,

learning

activities,

courseware

module
Bandura  (1997), Brusilovsky (1998), and Castells (2007) Collaborative filtering
Bandura (1997), Brusilovsky (2007), Content-based filtering
Brusilovsky (2007), Castells (2007), Chen (2005) Clustering, Knowledge 

discovery, metadata, Item repository theory,

Rule-based expert system, Artificial neural network

Adopted from (Ghauth & Abdullah 2009)

Nachmias (2003) mentioned that limited time factor hinders learners to proceed locating suitable learning information as they may end up with the unsuitable materials. Some researchers have identified these in recommender systems and adaptive hypermedia systems researches and proposed some solutions to overcome the problems. Although these technologies are considered to be personalized in delivering the learning materials to suit learners’ needs and preferences, yet some rooms for improvement still need to be fulfilled to ensure the materials provided are useful learning materials that par to learners’ quality standard.

Another significant school of thought; are they suit and settle accordingly to our National Philosophy of Malaysian Education? Some researchers may be interested to personalize the presentation of content, navigate links or course sequencing according to user’s preference in terms of using a user model. Others are motivated to use different approaches in a new e-learning recommender system for example; peer learning and social learning theory to ensure quality learning materials are adapted by learners and improved their learning style as well (Ghauth & Abdullah 2010).  Recently several methods have been developed for collaborative filtering of information (Resnick et al. 1994;  Shardanand & Maes 1995;  Breese et al. 1998). By using this approach, problems on locating the items that have same values of interests become more manageable.

2.1.1    What is World Wide Web and Its Evolution?

The World Wide Web (WWW) is often associated with the internet and this project allows link to be made to access information anytime in anywhere without boundaries. (Berners-Lee 1990). It has been going through a few stages of development from the PC era, and now moving on drastically towards what is known as the semantic web era. Figure 1 shows the stages of the web development; the pc era, the web 1.0, web 2.0, web 3.0 and web 4.0.

Figure 1: The Evolution of Web (source: http://hlwiki.slais.ubc.ca/index.php/Web_3.0)

The initial description of Web 1.0, Web 2.0 and Web 3.0 were based on the following list of Figure 2 (O’Reilly, 2005):

Web 1.0 Web 2.0 Web 3.0
Read-only era Read-Write-Publish era Read-Write-Collaborative era
45 million global users (1996) 1 billion+ global users (2006) “the portable personal web”
focused on companies focused on communities focused on the individual
home pages blogs life stream
owning content sharing content consolidating dynamic content
Britannica Online Wikipedia Metadata
HTML, portals XML, RSS widgets, drag & drop mashups
web forms web applications Semantic web
directories (taxonomy) tagging (“folksonomy”) user behaviour (“me-onomy”)
Netscape Google iGoogle, NetVibes
pages views cost per click user engagement
advertising word of mouth advertainment

Figure 2: The Features of Web 1.0, Web 2.0 and Web 3.0

  1. i. Web 1.0 – The Information Portal

O’Reilly (2005) refers the web 1.0 as the Information Portal before introduced the term Web 2.0 at the Media Web Conference in 2004. The time frame defined for this web (1990-2000). It is known as Read-Only era as it displayed static page. Web 1.0 portrayed like a huge library with lots of information to be absorbed. It used only HTML (Hypertext mark up language); basic language and consists of META Tags.

  1. ii. Web 2.0 – The Web as a Platform

Known as Read-Write-Publish era, web 2.0 is the easiest way for people to share their thoughts, data and communicate. It is user centric design sites and a platform for social networking for examples blogs, wikis; as “collaborative technology for organizing information on Web sites” (source Wikipedia) twitter, YouTube, RSS (Really Simple Syndication) is a relatively new technology for pushing edited content to the end user, Tagging ( a form of metadata used to describe web content), etc. Yet, still many things that make these applications inadequate to serve today’s computer user’s needs and the search engines are to be considered not yet intelligent enough to understand information like a human does.

  1. iii. Web 3.0 – Semantic and Intelligent Web

Web 3.0 represents an evolutionary shift in how people interact with the web, and vice versa. As it is, web 3.0 comprises three basic components: the Semantic Web, the Mobile Web, and the immersive Internet.

“People keep asking what Web 3.0 is. I think maybe when you’ve got an overlay of scalable vector graphics – everything rippling and folding and looking misty – on Web2.0 and access to a semantic Web integrated across a huge space of data, you’ll have

access to an unbelievable data resource.” (Berners-Lee, 2006).

Wheeler (2009) predicted the e-learning of web 3.0 is to have at least four key drivers:

  • Distributed computing
  • Extended smart mobile technology
  • Collaborative intelligent filtering
  • 3D visualisation interaction

He claimed that e-learning 3.0 will trigger away from the traditional institutions; far more in self-organised learning.  Why? Tools and services are easier to be accessed which lead to personalize our learning.  So to say, collaborative teaching and learning across distance will be much at east. In addition, he did mention that ‘artificial intelligent agents’ in software will enable computers to filter out whatever unwanted and allow what we do want. In short, Twitter can act as a network filter that is, it is still under control by the user even though it is built in with little intelligent software.

Twitter is claimed to be a big step forward into revolution of semantic predictive filtering. It filters as what Wheeler (2010) argued that twitter is an instant updated information of rich source where it is easy to be updated on widely global issues. Besides, it is such a personal learning network.  Choose properly, refined and reshaped, it is indeed can be effective for the users.  As it is, Wheeler (2009) claimed web 3.0 to be Read-Write-Collaborative web which may drag the users/ learners to unlimited boundaries globally in anytime, to be supported with intelligent solutions to search the web, document management and organize the content as well. Some might say that collaborative tools will exist.  3Dtool also, but not the school, since some of the teachers do not trust the kids with technology. Thus, the collaborative filtering for teaching in a learning 3.0 environment comes into play. Two aspects seem relevant for learning 3.0: psychological attitudes towards sharing and indexing/ navigating matter. For instance, twitter probably adds some more semantic content but needs to be indexed, but this still not available yet.

2.1.2    What is Web 3.0 – based Teaching and Learning?

Smith (2011) points that the future of learning with the web is now moving towards a better, smarter, faster, richer and even more complex learning environment.  Learning virtually using the internet and the websites is no longer the process where the information is given in a passive manner;   it is now becoming more active process. This is as mentioned by Bingham (2011),   the semantic web or also known as Web 3.0 is where the content will look for you rather than you actively seeking it as well as your activities and interests will determine what finds you, and it will be delivered on how you want it and to your preferred channel.

As noted in researches earlier, web 3.0 technologies (mobile learning, immersive technologies, and the semantic web are custom made for learning) offer higher information of acquisition as they are smart searchers with more targeted results that suit to the users/learners preferences and interact with the right content and subject matter experts. In fact, having people who are triggered on 3.0 environment, does correlate with successful, effective teaching and learning as well.

2.1.3    What is collaborative filtering?

Kangas (2002) stated collaborative filtering (CF) is an alternative method to rate “similar” users to predict the items that have not being rated. Basically, people with similar taste tend to like similar types. Thus, this can be the good indicator for the item to be rated personally. Thus, CF can improve in taking advantage of the information triggered as it compare of one’s likes and dislikes to others to predict preferences. For example, a moderator of a virtual discussion group like in Persatuan Siswazah Fakulti Pendidikan UKM (PSFPUKM)  acts as a “filter’’ in sharing the information among the peers. Yet, he/she has the control to filter out whatever unwanted and allow what the users want.    Other example is the Amozon.com where the registered users review books. Basically, Kangas (2002) did mention however, the mechanism behind collaborative filtering systems is the following:

  • a large group of people’s preferences are registered;
  • using a similarity metric, a subgroup of people is selected whose preferences are similar to the preferences of the person who seeks advice;
  • a (possibly weighted) average of the preferences for that subgroup is calculated;
  • the resulting preference function is used to recommend options on which the advice-seeker has expressed no personal opinion as yet. Hence, the focal point of CF is the collection of preferences among the users (cf. Shardanand & Maes 1995). This concept can be applied also in social filtering and the adaptive filtering as well.

3.0       The E-learning Recommender Systems and Adaptive Hypermedia Systems

3.1       The E-learning Recommender Systems

E-learning recommender systems provide users with personalized suggestions that suit to their preferences and interests. The system will select by matching to users profiles or user groups. As it is, the systems require “intelligent” interface to determine the interest of the user and use the input to make suggestion accordingly (Kangas 2002). Others like Osmar (2002) considers this “intelligent” software as a system that tries to propose actions to learners based on previous learners’ actions. Those recommendations could be in terms of simply a web resource, or any on-line activities such as reading online on messages posted on a conferencing system or perhaps doing an exercise posted on blogs and etc. As it, recommendation systems act as significant role in online reading in which learners are provided with guidance in searching and ranking references according to their preferences, knowledge bits, test items and etc (Rafaeli, et al. 2005).

3.2 Adaptive Hypermedia Systems

In this systems presentation and navigation assistance is personalized to fulfil the needs of the user. A user model is used to represent the user requirements and a content model is used to represent the content. By using a set of algorithms while interacting to the Adaptive hypermedia systems, user can select the most appropriate content to be presented (Bhosale 2006). This is supported by Brusilovsky (2001) claimed that the adaptive hypermedia is an alternative to the traditional of “one-size-fits-all” and it is built based on the goal, preferences and knowledge of each individual user, and adapt to the needs of the user by interacting with the systems throughout the process administered.

4.         Recommendations for Teaching and Learning in Malaysian Context

Based on the related literatures on the aspects of the web, the collaborative filtering, the recommender systems as well as the Adaptive Hypermedia Systems, some big question arises: Are these aspects fit into the Malaysian educational context? Are the teachers ready to implement in their teaching approach?  Some of the questions by Wheeler (2009) argued that;

  • Does e-learning 3.0 business-driven; but users will have their say whether to choose or leave it?
  • Are the teachers willing to give in to technology?
  • Are the students truly ready to be autonomous learners?
  • Children/ young learners can upgrade mindset to 3.0 learning environment. What about the adults? Did they downgrade? What made them do that? Parents? Educational system, non-employers, development of the brains……?
  • Do you think web 2.0 is a failure and is the course why web 3.0 is the bridge to fulfil the gap?

Again Wheeler (2010) pointed at the pedagogical level saying that where teaching and learning are concerned, Web Enhanced Learning (WEL) and other technology are the indicators that have potential in enhancing approaches to transform the quality of learning. Thus, WEL enables students to participate directly, in control and be responsible towards their own learning processes. They are now more engaged deeply and actively with their learning using WEL tools collaboratively. Hence, it seems a shame to account such phenomena to have such high expectation and activities of students within the social web which sometime contra to what being practiced within the classroom even among the teachers/ educators themselves.

Malaysian educational system is exam-oriented (The Star Online 2006 & Tun Hussin 2006). Teachers are too focusing on increasing the percentage of passes in the national examinations such as the UPSR, PMR, SPM and STPM that they rather willing to focus on the chalk and talk approach. Textbooks and only printed materials would be the main use since these are the things which are considered enough to make the students score with flying colours. Instead, Malaysians need a fresh and new philosophy in their approach to exams (Ahmad 2003).

5.0       Conclusion and Future Study

As users/ learners/ teachers/ educators or any laymen who have interest in technology, they need to be involved in a lot of speculation in the buzz of digital and education, the newness of technologies; whether we like it or not, we need to put lots of efforts into materialising the changing for teaching and learning to take place around the technology. It is high time to have rich conversation, conduct active teaching and learning research, but bear in mind the collaborative intelligent filtering need comes into play in order to ensure technology (3.0) won’t do any harm to the users/ learners especially to ensure the Malaysian National Philosophy to be materialised. Thus, as teachers/ educators to progress, we need to develop/ discuss/ scrutinise our own practice and make explicit pedagogies underpin the practice may meet the current demands. Upgrade our mindset with prominent features:

  • Curiosity
  • Eagerness to learn
  • Desire for collaborative
  • Embracing of change
  • Fondness of individual thinking; as a matter of fact is major tasks to be conducted by schools, parents and society as a whole!

References

Adomavicius, G.et al, (2005).“Toward the Next Generation of Recommender Systems: A Survey of the State-of-the-Art and Possible Exensions”, IEEE Transactions on Knowledge and Data Engineering, Vol. 17, No. 6, 2005, pp. 734-749.

Bandura, A. (1997)Social learning theory. Englewood Cliffs, New Jersey, Prentice-Hall.

Bhosale, D. (2006). AlcoZone: An Adaptive Hypermedia Based Personalized Alcohol Education. Blacksburg, Virginia.

Brusilovsky, P. (2001)“Adaptive Hypermedia”, User Modeling and User-Adapted Interaction, Vol.11, pp. 87-110.

Brusilovsky, P.et al (1998), “A Study of User Model Based Link Annotation in Educational Hypermedia”, Journal of Universal Computer Science, Vol. 4, No. 4,  pp. 429-448.

Brusilovsky, P. et al (2002), “From Adaptive Hypermedia to the Adaptive Web”, Communication of the ACM, Vol. 45, No. 5, 2002, pp. 31-33.

Castells, P.  et al (2007), “An Adaptation of the Vector-Space Model for Ontology-Based Information Retrieval”, IEEE Transaction on Knowledge and Data Engineering, vol. 19, no.2, special issue on Knowledge and Data Engineering in the Semantic Web Era,  pp. 261-272.

Chen, C. et al (2005), “Personalized E-Learning System Using Item Repository Theory”, Computers & Education, Vol. 44, No. 3, pp. 237-255.

Berners-Lee. T & Cailliau R, (1990). World Wide Web: Proposal for a HyperText Project. Available at: http://www.w3.org/Proposal.html

Berners-Lee et al., (2006). A framework for web science. Foundations and Trends in Web Science, 1(1).

Breese J.S., Heckerman D. and Kadie C. (1998), Empirical Analysis of Predictive Algorithms for Collaborative Filtering, Proceedings 14th Conference on Uncertainty in Artificial Intelligence, Madison WI: Morgan Kauffman.

B. Mobasher, et al., (2000).Integrating Web usage and content mining for more effective personalization,in E-Commerce and Web Technologies,” Lecture Notes in Computer Science (LNCS) 1875,S pringer-Verlag, September 2000.

Ghauth. &  Abdullah.  (2009). “Building an E-Learning Recommender System using Vector Space Model and Good Learners Average Rating”, 2009 Ninth IEEE International Conference on Advanced Learning Technologies. 978-0-7695-3711-5/09 $25.00 © 2009 IEEE

DOI 10.1109/ICALT.2009.161.

Ghauth. & Abdullah.  (2010). “An Empirical Evaluation Of Learner Performance In E-Learning Recommender Systems And An Adaptive Hypermedia System”, Malaysian Journal of Computer Science, Vol. 23(3), 2010 pp 141-152.

Gauth & Abdullah (2010).Measuring learner’s performance in e-learning recommender systems. Available at: http://www.ascilite.org.au/ajet/ajet26/ghauth.html, retrieved  2011-04-01.

Kangas, S. (2002). Collaborative filtering and recommendation systems.VTT Information Technology.Findland.

Available at: http://www.vtt.fi/inf/julkaisut/muut/2002/collaborativefiltering.pdf

Manolis, V. & Konstantinos, G. M. ( 2004). “Unison-CF: a multiple-component, adaptive

collaborative filtering system”. University of Macedonia, Dept. of Applied Informatics Parallel Distributed Processing Laboratory Egnatia 156, P.O. 1591, 54006, Thessaloniki, Greece.

Manos, P.,Ioannis, R., et al (2005).Incremental Collaborative Filtering for Highly-Scalable Recommendation algorithms, ISMIS. Work conducted at ICS-FORTH.

Mulpeter, D.(2009). The genesis and emergence of Web 3.0: a study in the integration of artificial intelligence and the semantic web in knowledge creation. Dublin Institute of Technology, brendan.tierney@dit.ie

Nachmias, R., & Segev, L. (2003). Students’ use of content in Web-supported academic courses. Internet and Higher Education, 6, 145-157.

Osmar R. Zaíane, (2002).”Building a Recommender Agent for e-Learning Systems,” icce, pp.55, 2002.International Conference on Computers in Education (ICCE’02).

O’Reilly 2005,What is Web2.0. Design patterns and business models for the next generation of software. O’Reilly Media, 2005.

Available at: www.oreillynet.com/pub/a/oreilly/tim/news/2005/09/30/what-isweb-20.html.

Resnick P, Iacovou N., Suchak M., Bergstrom, and Riedl J. (1994), ” GroupLens: An open architecture for collaborative filtering of netnews“, Proceedings of ACM 1994 Conference on Computer Supported Cooperative Work, Chapel Hill, NC: ACM, 175-186.

Shardanand U. and Maes (1995),  Social information filtering: Algorithms for automating “word of mouth”, Proceedings of CHI’95 — Human Factors in Computing Systems, 210-217

Smith,M.K(2011).Teaching in the Web 3.0 World. April4,2011   Posted in News, semantic, Technology.Available at: http://taxodiary.com/2011/04/teaching-in-the-web-3-0-world/

Shardanand U. and Maes (1995), Social information filtering: Algorithms for automating “word of mouth”, Proceedings of CHI’95 — Human Factors in Computing Systems, 210-217

Available at: http://pespmc1.vub.ac.be/collfilt.html

Bingham, T. (2011). ASTD (the American Society for Training & Development) and their article, “Better, Smarter, Faster : Web 3.0 and the Future of Learning.” Availableat:http://www.astd.org/TD/Archives/2011/Apr/Free/Apr11_Better+Smarter+Faster.htm

Wheeler, S.(2009).Learning with ‘e’: The edgless University?

Available at:http://steve-wheeler.blogspot.com/2009/06/edgeless-university.html.

Wheeler S. (2010). Teacher resistance to new technologies: How barriers to Web Enhanced Learning can be overcome. In G. Trentin and M. Repetto (Eds) Faculty Training for Web Enhanced Learning. New York, NY: Nova Science.

Wheeler S and Lambert-Heggs W (2010) Connecting Distance Learners and Their Mentors Using Blogs: The MentorBlog Project. Quarterly Review of Distance Education, 10 (4), 323-331.

Wikipedia Wiki , Wikipedia. Available at http://en.wikipedia.org/wiki/Wiki

White, D. (2005). Technical Approaches for the SPIRE project. JISC Repositories

Programme, Oxford University.

http://blog.thestar.com.my/permalink.asp?id=6291

http://pgoh13.com/exam_revamp.html http://pgoh13.com/exam_revamp.html

http://english.peopledaily.com.cn/200603/06/eng20060306_248163.html

http://www.igi-global.com/bookstore/article.aspx?titleid=1651