Mobile Recommendation Engine Using Contextual Event Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing item recommendation systems on mobile devices fail to consider local time and location information, and do not accurately determine the current environment of the user, leading to inappropriate recommendations, as they do not correlate these factors with real-world events or consider historical user behavior.
Innovation Solution
A system and method that generate recommendations on mobile devices by using a combination of time, location, venue, and event information, along with transactional history and behavior of other users, to determine current user interests and provide relevant recommendations for digital media, news, and physical or digital merchandise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing item recommendation systems are used, then recommendations are provided to users, but the recommendations are inappropriate because they do not consider local time and location information
Solution Approach 1:
The system proactively determines the user's current environment (location, time, venue, event) before generating recommendations, rather than waiting for user input. This preliminary environmental assessment ensures recommendations are contextually appropriate from the start
Solution Approach 2:
The recommendation system dynamically adapts to changing user environments by continuously monitoring location, time, and event data. As users move between different venues or events, the system adjusts recommendations in real-time to match the current context
2Productivity
If location-based services make recommendations using simple selection and sorting criteria, then recommendations are generated quickly, but they are not accurate because they do not consider historical user behavior or correlate with real-world events
Solution Approach 1:
The system merges multiple data sources including user historical behavior, current location, time, venue information, and real-world event data into a unified recommendation framework. This combination allows for both speed and accuracy by processing integrated information rather than separate factors
Solution Approach 2:
The system introduces an event database as an intermediary layer that correlates user location and time with real-world events. This mediator enriches basic location data with contextual event information, enabling more accurate recommendations without significantly increasing processing time
3Ease of operation
If the same venue is used for different events (sporting event one night, music performance another), then location-based recommendations can be generated, but they are inappropriate because the system does not consider the current event context
Solution Approach 1:
The system determines the specific event occurring at a venue before generating recommendations, rather than relying solely on venue type. This preliminary event identification ensures that recommendations match the actual context (e.g., sports merchandise at a stadium during a game, not concert merchandise)
Solution Approach 2:
The system changes the contextual parameters of recommendations based on the specific event detected at the user's location. The same physical location yields different recommendations depending on the event parameter (sporting event vs. music performance), making the system adaptable to venue multi-use scenarios
Data Source
AI summary
A system and a method generate a recommendation on a mobile device. The system and the method may use a time, a location, a venue and/or an event to generate the recommendation. Further, the system and the method may use an event database to determine current interests of the user. Still further, the system and the method for generating a recommendation on a mobile device may use a transactional history of the user and/or behavior of other users to generate the recommendation. The system and the method may recommend, for example, digital media, news and event information, editorial content and/or physical or digital merchandise. As a result, the system and the method may generate a recommendation that corresponds to the current interests of the user.