From "Relevance" to "Knowledge"

It not just finds data and information, but also discovers the connections and relationships between them, and forms organized knowledge and insights.
The widespread use of NLP and AI technologies makes search not only measure the relevance of a query and related documents, but also integrates semantics and knowledge into the search results.

From "General" to "Vertical"

It is hard to organize and understand the searched information and knowledge with existing keyword-based information retrieval approach, therefore deep learning and pre-training techniques were introduced into the IR area, bringing significant improvements in model effectiveness and transferability, paving the way for in-depth understanding of vertical domains.
Deep learning and large model technologies have significantly improved model performance and transferability, paving the way for in-depth understanding of vertical domain information and knowledge transfer across multiple industries at a low cost.

From "Information Retrieval" to "Decision Support"

From finding and presenting information and connections from a large amount of data, to combining data with a larger time span and space range along with reasoning and trend prediction.
The advancements in deep learning and large model technologies in the fields of inference and causal relationship discovery help individuals make better decisions and judgments.