Online shopping recommender system using hybrid method

As Internet usage becomes very common nowadays, including in Indonesia, people do their activities by relying on information gathered from world wide web. One of the examples is gathering information before buying product. Customers will search the product reviews before deciding whether to buy the products or not. Product reviews can be found by searching through search engine, reading on personal blogs, accessing from certain websites that provide reviews, or searching on electronic forum. Because searching on the net can be a daunting task, we propose a system that can provide product reviews, together with recommendations, focusing on Indonesia market. The recommendations will become a valuable resource for customers to narrow the search space on searching the specific products suitable for their needs. We took a sample on two groups with different characteristics. The focus of this research is to design good personal recommendations. Our system generates two types of recommendation: personal and item-based. User-based collaborative filtering is implemented to produce personal recommendation. And we implement item-based collaborative filtering on item's recommendation. The evaluation has been taken by conducting a survey on the users. The users should state whether they like the recommendation or not. The result shows that most of the users like the item-based recommendations. But the personal recommendation is only preferred by the user who already did some activities: rate or view an item.