Event Suggestion System Using Word Embedding for Personalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current recommendation systems for travel destinations and events lack effectiveness in personalizing suggestions based on user preferences, intent, and context, relying on outdated data and failing to adapt to real-time user inputs and changes in user behavior.
Innovation Solution
An event suggesting method utilizing word embedding to determine user preference information and intent, which involves assigning values to reference items from user input data, generating user intent information, and suggesting events based on similarity between user preference and intent information stored in a database, incorporating context updates and historical data for personalized recommendations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional recommendation systems use outdated data and simple algorithms, then system complexity is low, but recommendation accuracy and personalization are poor
Solution Approach 1:
The patent transforms user review data into vector representations using word embedding techniques, changing the parameter format from text to numerical vectors. This enables sophisticated similarity calculations and intent analysis while maintaining systematic processing capabilities.
Solution Approach 2:
The patent introduces an intent analysis module as an intermediary between user input and event recommendations. This mediator extracts user intent from reviews and uses it to filter and rank events, improving recommendation accuracy without requiring the entire system to be fundamentally complex.
2Adaptability or versatility
If the system processes diverse user inputs including text, pictures, and voice, then user satisfaction improves, but information processing complexity increases
Solution Approach 1:
The patent creates a universal processing framework that handles multiple input types (text, pictures, voice) through common pipelines. User reviews of any type are transformed into vector representations and processed through the same intent analysis and event matching mechanisms.
Solution Approach 2:
The patent extracts key features and intent from diverse input types, separating the essential information needed for recommendations from the raw data format. This extraction approach allows versatile input handling while keeping the core recommendation engine relatively simple.
3Reliability
If the system continuously updates user preference information based on real-time inputs, then recommendation relevance improves, but computational resource consumption increases
Solution Approach 1:
The patent pre-processes user reviews and calculates vector representations when reviews are submitted, storing these processed forms for future use. This preliminary action reduces the computational burden during real-time recommendation generation, as the heavy lifting of text-to-vector transformation has already been done.
Solution Approach 2:
The patent updates user preference information selectively based on relevant new inputs, rather than continuously reprocessing all data. The system focuses computational resources on processing new reviews that actually change user preferences, maintaining recommendation relevance while conserving resources.
4Measurement precision
If the system builds a comprehensive database from multiple user reviews and contexts, then recommendation personalization improves, but data management complexity increases
Solution Approach 1:
The patent transforms unstructured user review data into structured vector representations with defined dimensions and properties. This parameter transformation creates a standardized data format that simplifies storage, retrieval, and comparison operations in the database.
Solution Approach 2:
The patent segments user preference information into distinct components including user intent vectors, event vectors, and context information. This segmentation allows the database to organize and manage complex data through modular, manageable units that can be independently processed and combined.
Data Source
AI summary
Disclosed are a terminal, a server and event suggesting methods thereof, the event suggesting method using word embedding, the method including: determining user preference information about a predetermined object by assigning values to a plurality of reference items with respect to user input data corresponding to the object; generating user intent information, which includes values based on the plurality of reference items, based on at least one piece of review data corresponding to a plurality of events for recommendation objects, and accessing a database built up by adding the generated user intent information to the plurality of events; and suggesting an event following the object among the plurality of events based on a similarity between the user preference information about the object and the user intent information of each event included in the database. Thus, the event based on to a user's preference is suggested, thereby providing the event-recommendation service to make a user's satisfaction higher.