Knowledge mapping, like deep learning, is a key technology in artificial intelligence. Baidu CTO Wang Haifeng said, “Knowledge graph is the cornerstone for machines to understand the objective world like human beings.” Baidu BRAIN is the culmination of baidu AI’s technical accumulation and business practice for many years. Baidu Brain has been developing for nearly 10 years, and the knowledge graph technology was started in 2013. On the one hand, the scale of the knowledge graph is growing rapidly, and on the other hand, the magnitude of services based on the knowledge graph that can meet users’ various needs on a daily basis is also increasing rapidly. In the past few years, the size of Baidu’s Knowledge Graph service has grown about 490 times.
Artificial intelligence is divided into three stages, from machine intelligence to perceptual intelligence, and then to cognitive intelligence. Machine intelligence emphasizes more on the computing power of these machines, the processing power of large-scale clusters, and the processing power of gpus. On this basis there will be perceptual intelligence, perceptual intelligence is speech recognition, image recognition, from the picture to recognize a cat, face recognition, is perceptual intelligence. Perceptual intelligence is not unique to humans, but some of it can be found in animals. The next level of cognitive intelligence, which is unique to human beings, is based on thinking, and the establishment of cognition requires the ability to think, and thinking is based on knowledge, and you have to have the basis of knowledge, some common sense, to establish some thinking, to form a reasoning mechanism. AI needs to move from perceptual intelligence to cognitive intelligence. In essence, knowledge is a foundation, and reasoning based on knowledge happens to have such an attribute in the knowledge graph.
Knowledge graph is a simple database technology. The best open source knowledge graph technology in China is gStore, which exceeds Neo4j in many aspects. Knowledge graphs are much easier to learn than deep learning, but many online materials conceptualize and complicate knowledge graphs. Therefore, in order to enable you to quickly understand the knowledge Graph and its application scenarios, we will hold a technology salon of the knowledge Graph in Zhongguancun, Beijing on October 31st (Saturday).
Today is 1024 programmers Festival, we send out special benefits, limited time discount only today, 1 yuan registration, limited to 50 people to participate in the scene, need to sign up to the following poster introduction to share guests and themes, scan the QR code to sign up.
