This link opens in a new window
This link opens in a new window
This link opens in a new window
This link opens in a new window
Cover academic journals and academic collections across multiple disciplines
CNKI has launched AI-enhanced retrieval function. Welcome to try it out and provide feedback.
The trial period will last until December 31, 2024.
The "AI-enhanced retrieval" feature, building on the traditional retrieval service, integrates the natural language processing and semantic understanding capabilities of large models into information retrieval. It supports retrieving documents and document paragraphs using natural language and innovates from traditional keyword-based retrieval to semantic vector-based retrieval. It also refines the granularity of retrieval from document retrieval to paragraph retrieval, upgrades the retrieval experience from simple retrieval to intelligent interaction, and enhances the quality of service from literal retrieval to standardized guidance retrieval. It also provides more intelligent applications such as generating citations and reading paragraphs of the same topic. It greatly improves the quality and efficiency of literature research, academic innovation, professional retrieval, and evaluation.
Three retrieval methods are provided:
The "AI-enhanced retrieval" feature, building on the traditional retrieval service, integrates the natural language processing and semantic understanding capabilities of large models into information retrieval. It supports retrieving documents and document paragraphs using natural language and innovates from traditional keyword-based retrieval to semantic vector-based retrieval. It also refines the granularity of retrieval from document retrieval to paragraph retrieval, upgrades the retrieval experience from simple retrieval to intelligent interaction, and enhances the quality of service from literal retrieval to standardized guidance retrieval. It also provides more intelligent applications such as generating citations and reading paragraphs of the same topic. It greatly improves the quality and efficiency of literature research, academic innovation, professional retrieval, and evaluation.
Three retrieval methods are provided:
- Quick retrieval: Supports input of both short and long descriptions, multiple keywords, retrieval formulas, and different languages.
- Advanced retrieval: Retains traditional advanced retrieval items, supports complex logical combinations, and allows for more complete expressions including technical terms, application fields, research methods, and research objectives.
- Paragraph retrieval: Supports input of short and long descriptions, multiple keyword inputs, and conceptual queries. It retrieves directly to the original document paragraph and can compare and trace the original document paragraph of the same topic, while also generating intelligent citation text.
It also provides more powerful content input and result output:
- Input Enhancement: Supports natural language and voice input; intelligently recognizes retrieval intent, generates extended retrieval, and expands and adjusts input expressions; intelligently prompts retrieval words and guides standardized retrieval.
- Result Enhancement: Balances accuracy and completeness, recalls two ways, integrates semantic and keyword retrieval, and combines input models such as query characteristics, relevance characteristics, document quality characteristics, and historical click characteristics; stronger semantic related recall ability, cross-language recall ability, and fault tolerance ability; supports more standardized retrieval such as authors, institutions, and publications."
Harness the power of comparable data & insights.
Data Science Studio
This starting point offers a consistent process for building and training machine learning models, regardless of the type of problem being solved (regression, classification, or text classification). It includes a range of algorithms and libraries, including scikit-learn and TensorFlow, that can be used to build and train machine learning models for these different types of problems.
GlobalData Explorer → Tools → Data ScienceStudio
Key capabilities: