General-purpose vs. Industry Algorithm Edition for education

Industry Algorithm Edition for education

To address challenges in question-answering scenarios, the Industry Algorithm Edition for education is built on the latest algorithms. It offers features tailored for the education industry, such as intelligent semantic understanding for question-answering, vector retrieval, and ranking algorithms. These features ensure high performance and result accuracy for search services. This edition resolves key industry challenges, including high search latency from large question banks and high zero-result rates from incomplete ones. It provides a multimodal search solution. For the education industry, OpenSearch also provides a vector model to enable vector retrieval and multimodal search, which improves question-search accuracy. For more information, see multimodal search.

Feature comparison

Feature	General-purpose Edition	Industry Algorithm Edition for education
One-stop configuration	After you create an application, you must manually create and configure query analysis, a sort policy, and a drop-down suggestion model.	Tailored for common education search scenarios, this edition provides application schema and index schema templates for one-click configuration, which simplifies setup for new users.
Query analysis	Offers query analysis features for general industries, including synonym configuration, stop word filtering, spelling correction, term weight analysis, and category prediction.	Provides an enhanced analyzer and query analysis features for general and question-answering scenarios in the education industry. By addressing specific use cases and challenges, it builds more precise indexes, better understands user query intent, and delivers superior performance compared to the General-purpose Edition.
Sort policy	You must manually configure and tune the sort policy for your application based on business scenarios.	Built on the application schema and index schema templates, this edition provides common baseline and business-specific ranking expressions for the education industry. This meets most ranking requirements without requiring additional configuration.
Feature iteration	Regularly updates system default dictionaries for components such as the analyzer and query analysis.	Continuously updated to reflect changes in terminology and products within the education industry. It optimizes existing tokenization and query analysis capabilities to provide more timely and effective search results.

Query analysis performance comparison

Compared with the General-purpose Edition, the Industry Algorithm Edition for education provides deeper optimization in query analysis. In addition to correcting common errors in the general-purpose dictionaries, it incorporates best practices from across the education industry to enrich the existing dictionaries. The following examples show these improvements:

• Tokenization (space-based)

Query	General-purpose Edition	Industry Algorithm Edition for education
Atteno's clockwehaveourbreak.	Atteno ' s clockwehaveourbreak .	At ten o'clock we have our break .
BeforeIgottonthebusstop,thebushadalreadyleft.	BeforeIgottonthebusstop , thebushadalreadyleft .	Before I got to the bus stop , the bus had already left .
Itis20yuan.	Itis 20 yuan .	It is 20 yuan .
minusfrac12xmsup3yplus3xymsup3minus5x	minusfrac 12 xmsup 3 yplus 3 xymsup 3 minus 5 x	minus frac 12 x msup 3 y plus 3 xy msup 3 minus 5 x
adequal2,bcequal4	adequal 2 , bcequal 4	ad equal 2 , bc equal 4
Vertex leftoright of rectangle leftoabcright	The vertex leftoright of the rectangle leftoabcright	The vertex left o right of the rectangle left oabc right

Usage notes

• To create an Industry Algorithm Edition for education application, see Industry Algorithm Edition for education.

• An exclusive General-purpose Edition application can be converted to an Industry Algorithm Edition for education, but an Industry Algorithm Edition for education application cannot be converted back to the General-purpose Edition.

• The Industry Algorithm Edition for education applies only to exclusive applications.

• If you upgrade a shared application to an exclusive one and then adapt it for the education industry, ensure that the specifications of the online application and the instance are both exclusive before you perform the Industry Algorithm Edition for education adaptation.

On the application schema configuration page, the field tag column displays the tag for each field. Ensure that each field tag is mapped to the corresponding field name. Otherwise, an error will occur when you save the configuration.

The index schema configuration page contains three sections: the Index Field List (which shows the index name and tokenization type for each field), the Attribute Field List (for fields that can be used in filter, aggregate, sort, and distinct clauses), and the Default Display Field List. After confirming that each index tag is correctly mapped to an index name, click Next.