Shulou(Shulou.com)06/03 Report--

This article mainly introduces "what is the knowledge map". In daily operation, I believe many people have doubts about what is the knowledge map. I have consulted all kinds of materials and sorted out simple and easy operation methods. I hope it will help you answer the doubts about "what is the knowledge map"! Next, please follow the small series to learn together!

01 What is Knowledge Graph?

We can look at the concept of knowledge graphs from different perspectives.

From a Web perspective, knowledge graphs support semantic search by establishing semantic links between data, just like hyperlinks between simple texts.

From the perspective of natural language processing, knowledge graph is to extract semantic and structured data from text.

From the perspective of knowledge representation, knowledge graph is a method of representing and processing knowledge by computer symbols.

From the perspective of artificial intelligence, knowledge graph is a tool that uses knowledge base to assist in understanding human language.

From the database perspective, knowledge graph is a way to store knowledge by using graph.

At present, the academic community has not given a unified definition of knowledge graph. Explicitly described in Google's published documentation, Knowledge Graph is a technical way to describe knowledge and model the relationships between everything in the world using graphical models.

Knowledge graph is also a more general formal description framework of semantic knowledge. It uses nodes to represent semantic symbols and edges to represent relationships between semantics, as shown in Figure 3-1. In knowledge graphs, people, things, and things are often referred to as entities or ontologies.

▲ Figure 3-1 Knowledge Map Example

The knowledge graph consists of three elements: entities, relationships and attributes.

Entity: also known as ontology, refers to objective existence and can be distinguished from each other things, can be specific people, things, things, but also abstract concepts or connections. Entities are the most basic elements of the knowledge graph.

Relationships: In the knowledge graph, edges represent relationships in the knowledge graph, used to represent some relationship between different entities. Figure 3.1 illustrates the relationship between Turing and AI, the relationship between the Knowledge Graph and Google, and the relationship between Google and deep learning.

Attributes: Entities and relationships in the knowledge graph can have their own attributes, as shown in Figure 3-2.

▲ Figure 3-2 Attributes in Knowledge Graph

Knowledge graph construction involves knowledge modeling, relationship extraction, graph storage, relationship reasoning, entity fusion and other aspects of technology. The application of knowledge graph is embodied in semantic search, intelligent question answering, language understanding, decision analysis and other fields.

02 The Value of Knowledge Graph

Knowledge graph was first applied to search engine. On the one hand, concept retrieval was realized through reasoning, on the other hand, structured knowledge classified and arranged was displayed to users in a graphical way, so that people could be freed from the mode of searching for answers by artificial filtering web pages, and it could be applied to intelligent question answering, natural language understanding, recommendation and so on.

The development of knowledge graph benefits from the development of Web technology, influenced by KR, NLP, Web and AI. The value of knowledge graphs comes down to making AI smarter.

1. Help Search

The purpose of search is to enable people to find something easily and quickly in the Internet of Everything. At present, our search habits and search behavior are still based on keywords for search purposes, and the emergence of knowledge graphs can completely change this search behavior pattern.

When the knowledge graph has not been applied to the search engine, the search process is to find the URL with the highest matching degree with the query from the mass URL, and return some results with the highest ranking score to the user according to the query results. In the whole process, search engines may not need to know what users input, because the system does not have the ability to reason, and it is also slightly insufficient in accurate search.

Knowledge graph-based search, in addition to being able to answer users 'questions directly, also has certain semantic reasoning ability, which greatly improves the accuracy of search. Figure 3-3 shows a schematic diagram of knowledge graph assisted search.

▲ Figure 3-3 Knowledge Map Helps Search

2. Help recommendation

Recommendation technology and search technology are very similar, but slightly different. Search technology uses information pull, while recommendation technology uses information push, so there are some problems in recommendation technology, such as cold start and data sparse.

Take e-commerce recommendation as an example to introduce the application of knowledge graph in recommendation. Suppose I buy a mobile phone, and the strong subordinate relationship of the mobile phone is phone case, so that the system can recommend the phone case to me, and it can also recommend similar or complementary entities. Figure 3-4 is a schematic diagram of knowledge graph assistance recommendation.

▲ Figure 3-4 Knowledge Map Helps Recommendation

3. Help Questions

Question answering and dialogue system has been one of the key symbols of NLP in the field of artificial intelligence implementation. The knowledge graph is equivalent to loading a background knowledge base for the question and answer and dialogue system.

For question-answering and dialogue systems or chatbots, in addition to sparse large graphs in open fields such as entity knowledge graphs and interest knowledge graphs, dense small graphs for robots and users are also needed. At the same time, knowledge graphs need to be updated dynamically. Figure 3-5 is a schematic diagram of knowledge graph assisted question answering.

▲ Figure 3-5 Knowledge Map Helps Ask Questions

03 Knowledge Graph Architecture

The architecture of knowledge graph involves knowledge representation, knowledge acquisition, knowledge processing and knowledge utilization.

In general, the knowledge graph construction process is as follows: first determine the knowledge representation model, then select different knowledge acquisition methods according to different data sources and import relevant knowledge, then use knowledge reasoning, knowledge fusion, knowledge mining and other technologies to construct corresponding knowledge graphs, and finally design the representation mode of knowledge graphs according to different application scenarios, such as semantic search, intelligent recommendation, intelligent question answering, etc.

Logically, we can divide the knowledge graph into two layers: the data layer and the schema layer. The data layer can be a database stored in units of facts, and the graph databases that can be selected include RDF4j, Virtuoso, Neo4j and other triples.

Or it can be stored in a graph database as a basic representation. The schema layer is built on top of the data layer and is the core of the knowledge graph. Typically, the data layer is managed through an ontology library, the concept of which is equivalent to the concept of "class" in an object. With the help of ontology library, we can manage axioms, rules and constraints, standardize the relationship between entities, relationships and attributes.

Knowledge graphs can be constructed from top to bottom and from bottom to top. Top-down construction refers to extracting ontology and schema information from encyclopedic data sources and adding them to knowledge base. Bottom-up construction refers to extracting resources from open data with the help of certain technical means, selecting information with high confidence, and adding it to the knowledge base after manual review.

At the beginning of the development of knowledge graph, most enterprises and institutions adopt top-down approach to construct knowledge graph. At present, most enterprises adopt bottom-up approach to construct knowledge graph.

The architecture of the knowledge graph is shown in Figure 3-6.

▲ Figure 3-6: Structure of Knowledge Graph

Knowledge sources: including structured data, unstructured data and semi-structured data.

Information extraction: extracting entities, attributes and relationships among entities from various types of data sources, and forming ontological knowledge representations on this basis. There are a lot of unstructured or semi-structured data in the process of knowledge graph construction, which need to be extracted by natural language processing. From this data, we can extract entities, relationships, and attributes.

Knowledge fusion: The main work is to extract entity information from structured data and information, and even to align and disambiguate entities from third-party knowledge bases. The output of this phase should be the fusion of ontology information from various data sources.

Knowledge processing: The knowledge processing stage is shown in Figure 3-6, in which the important work in knowledge reasoning is the completion of knowledge maps. Common completion methods of knowledge graph include: completion method based on ontology reasoning, related reasoning mechanism implementation and completion method based on graph structure and relationship path characteristics.

At this point, the study of "what is knowledge map" is over, hoping to solve everyone's doubts. Theory and practice can better match to help you learn, go and try it! If you want to continue learning more relevant knowledge, please continue to pay attention to the website, Xiaobian will continue to strive to bring more practical articles for everyone!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.