System architecture of a ubiquitous web-wide search engine

The position of search engine in information world is to fill the information fault of people and information world, and the big search service technology oriented to ubiquitous network is to combine people, things and information organically to provide users with intelligent service and solution. Internet search engine is only for the text, and in the future market demand, there must be in the IoT, ubiquitous in the web search for a variety of objects, relationships, knowledge of intellectual needs. In the Internet, search engine is the internet entrance, who mastered the search engine technology, who in the Internet application market has made a favorable position. Similarly, in the ubiquitous network, who mastered the ubiquitous web search engine technology at the core, who in the ubiquitous network, the Internet of things applications in the market, there will be a huge advantage.

The large-size search in the Web mainly includes 5 aspects of content, Universal Data acquisition, Multi-source fusion knowledge acquisition, user Search intention understanding, Intelligent Solution Online matching and security privacy protection.

is a ubiquitous web search engine architecture, combined with this diagram, the role of the 5 modules to explain.

First, the Universal Data acquisition , that is, to support the access to the ubiquitous network space data and information, including the Internet, IoT, social networks, medical health, video surveillance, geographical information and other space to collect a variety of social resources, information resources, physical resources data. Data acquisition and acquisition of large-size search is purposeful around the solution to collect data, including grammatical and semantic related data. In the data type aspect, including the text, the picture, the speech, the video and so on various kinds of multimode data.

Secondly, multi-source fusion knowledge acquisition , that is, based on the universal network to obtain data, facing the vast number of entities in cyberspace and the relationship into the knowledge mining, through integration, correlation, statistics, reasoning, and even crowdsourcing methods to support large-scale entities and relationships and other knowledge, timely empty attribute representation and acquisition Knowledge aggregation in the knowledge is supported after two processing, through the user's query, modification, feedback and self-evolution process, the process of gradual improvement, the process is the core process of the big search engine.

Thirdly, the user search intention understanding , that is, the basic needs of the accuracy and ambiguity elimination of the intention understanding, combined with the user's context and semantic knowledge and other methods, quickly and accurately understand the user's real intentions, and transformed into a knowledge aggregation can match the presentation of the deduction. Big Search is a new generation with "wisdom" of the search, the user can query input keywords, voice, gestures and other content, combined with the user mobile phone terminal, the space-time scene of the motion trajectory and historical records and personal preferences and other information, accurate understanding of the user's intentions, and the use of a unified model to support efficient query deduction is expressed.

Finally, the online matching of wisdom solution is based on the knowledge aggregation and index after intention understanding representation and index, through fast matching, sequencing and other techniques, to form several intelligent comprehensive solutions satisfying users ' real intentions, and to give their relevance sort through the result evaluation method. In large search, search is a kind of intelligent service, knowledge deduction process is based on the accurate understanding of the needs of users, in an optional range, for users to quickly provide intelligent solutions.

Security privacy Protection, that is, to ensure that the entire user search is safe. The first is to ensure that the data source and processing results are credible, followed by mining search for the user's privacy is not exposed and malicious use, while supporting the violence, pornography and other malicious information filtering. In the process of data collection and acquisition, the data source is confirmed to ensure the traceability of the information, and finally the result of the search is credible. Large search and fine-grained access control tests ensure that the right search results are returned only to the right user without being abused, and that the user's privacy is not compromised or maliciously analyzed using relevant technical means.

These five aspects are ubiquitous web search engine design a kind of idea. Domestic and foreign academic circles have certain research in these five aspects, and each face a number of challenges need to be solved. I think it is a very meaningful thing to study these problems and how to combine them organically. The Internet of things market is a very promising market, mastering the Internet of Things search technology, undoubtedly to seize the entrance of this market, has a huge advantage.

System architecture of a ubiquitous web-wide search engine