A summary of the Knowledge Atlas (knowledge graph)

A summary of the Knowledge Atlas (Knowledge graph)

The main purpose of the Knowledge Atlas (knowledge graph) is to display the knowledge to the user in a more intuitive way in the mass network data.

Two, KG features and functions: essentially a semantic network, the user Query keyword mapping to the concept of semantic knowledge base,

The core is the knowledge base.

Knowledge map process: Knowledge acquisition, knowledge fusion, knowledge storage, query-type semantic understanding, knowledge retrieval and visualization

3.1 Knowledge Acquisition:

In the big Data Environment, through machine learning, knowledge mining, natural language processing and other ways to obtain, this shows that in learning the knowledge map, for machine learning, data mining and other related directions should also have a certain understanding and use.

3.2 Knowledge Fusion:

In the face of the large amount of data obtained on the network, its format is equal to unify, which is called heterogeneous data.

Entities and relationships are the basic elements of a map, and as with vertices and edges in the graph, the importance of the entity is based on

Sorting by algorithms such as PageRank

In the process of knowledge fusion, it is necessary to use inference to detect logical contradiction and integrate knowledge into concrete and classify by aggregation and classification.

Because any entity is not isolated, and some of the concepts around a certain logical relationship, it can be seen in the discrete mathematics of mathematical logic, relationship proof and other basic knowledge of the importance, and the combination of graph theory.

3.3 Knowledge Storage:

Relies heavily on massive data storage technologies to manage large-scale distributed data for high capacity, scalability, reliability, and performance.

3.4 Query-Type semantic understanding:

Various query methods, query-based semantic analysis, word segmentation, labeling, error correction, grammar analysis, and knowledge base matching, user sentiment and context, query expansion, mainly in natural language processing and artificial intelligence based.

3.5 Knowledge Search:

This stage involves information retrieval, knowledge Mining Technician key technology (similarity, importance). The query semantics are parsed, the knowledge of knowledge base is matched, and the statistics, sequencing, inference and prediction are carried out. Provide users with critical, complete and accurate information, and recommend information that may be of interest to the user (referring to the referral system).

3.6 Visual Presentation:

Enhance user experience and effect, more attention to content display granularity on the grasp, need to involve Web client technology, visualization technology, human-computer interaction technology to help users improve the experience.

Iv. application of the Knowledge Atlas

4.1 Commercial Search engine applications: such as Baidu, Sogou, etc., foreign Google.

Application of the 4.2 question and answer system: Apple's Siri

4.3 E-commerce platform application: Taobao

4.4 Social Network usage: FB

4.5 Other areas: such as educational research, medical, biological research

Five Summary:

As a new knowledge organization and retrieval technology in the Big Data era, the Knowledge Atlas has gradually manifested the advantages of knowledge organization and display, and has been paid attention to in many fields. The application foreground is very wide. At present, the development of Knowledge Atlas is still in the initial stage, facing many challenges and problems, such as: automatic extension of knowledge base, heterogeneous knowledge processing, inference rule learning, cross-language retrieval and so on.