preview

Personalized Tag Recommendation For Support The Future Learning Needs Of Learners

Decent Essays

5 PERSONALIZED TAG RECOMMENDATION
In personalizing recommendation to support the future learning needs of learners, we used hybrid recommendation approach whereby recommendations are based collaborative filtering (learning interest of other community members) and content-based filtering (previous learning interest of individual learner). This section describes the procedure taken in predicting the future learning needs for each learner. Also, in this section we examined the long and short term implications of using long and short term learning data in predicting future gaps in the knowledge of learners.

5.1 Inferring User’s Learning Interest
In order to personalize learning for each user we seek to first determine their indi-vidual learning interest by determining the tag class where their interest lies and there-after determining the tag distributions represented in their learner model to determine their specific learning interest. In eliciting the learning interest of individual learner, we inferred their long term and their short term learning interest. Questions asked from January 2009 to December 2011 were used to infer their long term learning needs while questions asked within January 2014 to July 2014 were used to depict the short term learning needs of the user. The specific learning interest of the learner in both long and short term was determined by mining all tags employed in questions asked by the user in the long and short term time period. In determining

Get Access