Virtual goods recommendations in virtual worlds.

Area of Science:

• Computer Science
• Human-Computer Interaction
• Social Computing

Background:

• Virtual worlds (VWs) are computer-simulated environments where users interact via avatars.
• Rapid growth in VWs leads to hasty virtual goods releases, often resulting in irrelevant items.
• This mismatch hinders remarketing and user satisfaction, impacting virtual economy efficiency.

Purpose of the Study:

• To propose a novel virtual goods recommendation method for VWs.
• To leverage social interactions within VWs to improve recommendation accuracy.
• To address the challenge of irrelevant virtual goods and enhance user experience.

Main Methods:

• Developed a recommendation system incorporating social interaction data.
• Quantified social interaction strength and influence among users.
• Utilized data from an online virtual world platform for experimentation.

Main Results:

• The proposed method, considering social interactions and social circles, demonstrated superior performance.
• Social neighbor interactions significantly influence users' virtual goods purchasing intentions.
• The system effectively improved the relevance of recommended virtual goods.

Conclusions:

• Social interactions are crucial for effective virtual goods recommendation in VWs.
• Integrating social dynamics enhances user experience and purchasing decisions.
• This approach offers a more personalized and efficient recommendation strategy for virtual environments.