Negative Training Set Generation via Search Session Logs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine learning algorithms face challenges in generating effective negative training examples, which are crucial for training supervised learning models but are often difficult to create accurately.
Innovation Solution
A computer-implemented method and system for generating a training set for machine learning algorithms by retrieving queries and search engine result pages from a search log database, identifying search sessions, and creating negative training examples by pairing a second query with a predetermined search result from the first query's results set, ensuring the queries are submitted within the same search session.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If random negative examples are used for training, then the training process is simple, but the training effectiveness is poor
Solution Approach 1:
The system uses the search engine's own query logs and search sessions to automatically generate negative training examples. The search engine serves itself by utilizing its existing operational data (queries, SERPs, user interactions) to create training data, eliminating the need for external manual annotation while improving training effectiveness through contextually relevant negative examples.
2Manufacturing precision
If manual annotation is used to create negative examples, then the training data quality is high, but the time and resource consumption is excessive
Solution Approach 1:
The system copies existing search data structures (queries, SERPs, user interaction patterns) from the search engine's operational logs to create training examples. By replicating and reusing existing high-quality search data rather than manually annotating new data, the system maintains training data quality while dramatically reducing the time and resources required for training set creation.
3Reliability
If contextually relevant negative examples are generated using search session analysis, then the training effectiveness is improved, but the system complexity increases
Solution Approach 1:
The system analyzes user interaction feedback (clicks, dwell time, navigation patterns) within search sessions to automatically identify and generate negative examples. This feedback mechanism allows the system to learn from actual user behavior patterns, improving training effectiveness by creating negative examples that reflect real search scenarios without requiring complex manual intervention.
Data Source
AI summary
There is disclosed a method and system for generating a training set for training a machine learning algorithm (MLA) implemented in an information retrieval system. The method is executable by the server and comprises: retrieving, from a search log database of the server, a first query previously submitted to the server, a first SERP associated with the first query, a second query different from the first query and submitted after the first query, and a second SERP associated with the second query, the first query and the second query having been submitted by the electronic device: the first SERP including a first set of search results; and the second SERP including a second set of search results; in response to the second query being submitted within a same search session as the first query, generating the training set to be used as negative training examples for training the MLA.