Beijing Tulex Information Technology Co., Ltd. General Manager is water
Currently, information retrieval has evolved to network and intelligent phases. The object of information retrieval is relatively closed, stable, and the information content managed by the independent database is extended to open, dynamic, updated fast, widely distributed, manage loose web content; information retrieval is extended by the original intelligence professionals. Including business people, managers, teachers, various professionals, etc., they have put forward higher and more sample requirements from the results from the results to the way. Adapt to networking, intelligence, and personalization needs are new trends in the development of information retrieval technology. The hotspot intelligent retrieval or knowledge of information retrieval technology is based on keyword matching. It is often found that there is a phenomenon that is insufficient, not allowed, and the quality of retrieval quality is not high, especially in the network information age, using keyword matching It is difficult to meet the requirements of people's retrieval. Intelligent retrieval utilization word dictionary, synonymous dictionary, homonyl dictionary improves retrieval effect, such as user query "computer", information related to "computer" can also retrieve it; further can assist in inquiry on knowledge level or conceptual level, pass The theme dictionary, the upper and lower dictionary, related class dictionary, form a knowledge system or concept network, give users intelligent knowledge tips, ultimately help users get the best search effect, such as users can further reduce queries to "microcomputer", "server "Or expand the" electronic technology "," software "," computer application "such as" Information Technology "or inquiry-related" Electronic Technology "," Computer Application ". In addition, intelligent retrieval also includes ambiguity information and retrieval processing, such as "Apple", refers to fruit or computer brands, "Chinese" and "People's Republic" distinctive, will pass ambiguous knowledge description library, full-text index, user retrieval context Analysis and user-relevance feedback techniques combined with processing, efficient and accurate feedback to the user's most needed information. Knowledge mining currently mainly refers to the development of text mining technology to help people find better discovery, organization, representation information, extraction knowledge, meet the high level of information retrieval. Knowledge mining includes summary, classification (clustering) and similarity retrieval. The automatic summary is to automatically extract abstracts from the original literature using the computer. In the information search, the automatic summary helps the user quickly evaluates the correlation degree of retrieval results. In the information service, automatic summary helps a variety of forms of content, such as sending to PDA, mobile phones, etc. Similarity Retrieval Technology Retrieves Document Content Features Searching with its similar or related document, is the basis for realizing user personalized related feedback, or for dearing analysis. Automated classification can form a predefined classification tree based on statistics or rules, and then classify it according to the content features of the document; automatic clustering is based on the correlation degree of document content and. Automated classification (clustering) is very useful in information organizations and navigation. Heterogeneous information integrated retrieval and hologram retrieval under information retrieval distribution and networkization, the openness and integration requirements of information retrieval systems are getting higher and higher, requiring information on different sources and structures, which is different. Structure information retrieval technology development base, including supporting various formatted documents, such as Text, HTML, XML, RTF, MS Office, PDF, PS2 / PS, MARC, ISO2709, etc., support multilingual information retrieval; support Unified processing of structured data, semi-structural data, and unstructured data; and seamless integration of relational databases and other open retrieval interfaces. The concept of "holographic search" is to support all formats and methods. From now on, it is still waiting for human-computer interaction and multimedia information retrieval integration based on natural language. Further breakthrough. In addition, from the perspective of engineering practice, multi-level caching, distributed clustering and load balancing technology, distributed cluster, and load balancing technology, which use memory and external storage, is also an important aspect of the development of information retrieval technology. With the popularity of the Internet and the development of e-commerce, enterprises and individuals can get it, the amount of information to be treated is explosive growth, and most of them are unstructured and semi-structural data.
