Holistic system for storage and transformation of information based on associative model of data. Целостная система для хранения и обработки информации, основанная на ассоциативной модели данных.
Search for text part, concrete object or anything using continuous qualification of request.
For example:
C# json
parsing json using C#
generate C# classes based on json
...
All possible actions or more precise requests can be collected from user requests and also generated by the search engine (before any search or on demand). It will be great if the search engine publishes all its search requests, so other systems can do fewer requests to that search engine directly.
This also can be used to search objects in database, each word will be a restriction of one of the attributes of the object (on of the collection of objects with the same type/structure) we are looking for.
The key idea is to compose a precise request before giving any results. This will look like narrowing matched results step by step. It is also similar to communication between people. A person can ask more questions before actually answering.
The system should prioritize such questions that can narrow the amount of search results in half. So it will be possible to ask few questions and give exact single matching search result at the end of Q and A session.
It also might be helpful to use a few questions to define the exact search target: is it an image, a page, a product or some other object type?
Konard
commented
3 years ago
Search for text part, concrete object or anything using continuous qualification of request.
For example:
C# json
All possible actions or more precise requests can be collected from user requests and also generated by the search engine (before any search or on demand). It will be great if the search engine publishes all its search requests, so other systems can do fewer requests to that search engine directly.
This also can be used to search objects in database, each word will be a restriction of one of the attributes of the object (on of the collection of objects with the same type/structure) we are looking for.
The key idea is to compose a precise request before giving any results. This will look like narrowing matched results step by step. It is also similar to communication between people. A person can ask more questions before actually answering.
The system should prioritize such questions that can narrow the amount of search results in half. So it will be possible to ask few questions and give exact single matching search result at the end of Q and A session.
It also might be helpful to use a few questions to define the exact search target: is it an image, a page, a product or some other object type?