32results about How to "Comprehensive search results" patented technology

The invention provides a full-text database accurate and efficient retrieval method for perfecting subject terms. The method comprises the steps of firstly, creating a long-tail subject term lexicon,designing a long-tail subject term selection algorithm, selecting a user to optimize a search expression, and clearing the search intention; designing a subject term synonymy expansion algorithm basedon the word forest expansion version, and expanding the subject term into a synonymy set of the subject term; and designing an improved synonym similarity calculation method based on the Known synonym tree, and calculating the similarity between the expanded subject term set and the original subject term by adopting an improved algorithm; setting an appropriate similarity threshold value throughmultiple groups of experimental statistical analysis, and simplifying the subject term set according to the similarity critical value to obtain an improved subject term set; and finally, improving a retrieval result sorting algorithm. The recall ratio, the precision ratio and the retrieval efficiency of the method are greatly improved, and good performance is shown.

The invention discloses a novel method for domain knowledge representation and service, namely, a knowledge banyan net representation system and a service framework. The knowledge banyan net representation system can represent declarative knowledge and procedural knowledge simultaneously, the declarative knowledge described by using the knowledge banyan net representation system contains attribute description knowledge and relevance knowledge, and the procedural knowledge refers to inferential rules and methods on the basis of the declarative knowledge described by using the knowledge banyan net representation system. A domain system and a knowledge service platform can be quickly established for providing management and retrieval of data, information and knowledge for a domain by using the knowledge banyan net representation system and the service framework. A knowledge representation system has the characteristics of little domain restriction and various representation forms, further supports sharing among different domain knowledge, and provides a flexible and intelligent method of knowledge representation and resource management and retrieval for vertical domain searching.

The invention relates to the field of question and answer retrieval and the blockchain technology, and particularly discloses an intelligent question and answer method and device, equipment and a storage medium. The method comprises the following steps: acquiring a to-be-retrieved question; determining a keyword of the to-be-retrieved question, and determining a first retrieval result according to the keyword based on a first retrieval model; inputting the to-be-retrieved problem into a pre-trained second retrieval model to obtain a second retrieval result; performing fusion sorting on the first retrieval result and the second retrieval result to obtain a sorting result; and determining a question and answer result corresponding to the to-be-retrieved question according to the sorting result, and feeding back the question and answer result to the user. Therefore, the answering accuracy of the question-answering system is improved. The first retrieval result, the second retrieval result and the sorting result can be stored in the block chain.

The embodiment of the invention provides a data query method and device, computer equipment and a storage medium, and belongs to the technical field of artificial intelligence. The data query method comprises the steps that a data query request is obtained, a historical browsing record of a target browser in a preset time period is obtained according to a data query instruction, the historical browsing record comprises a historical access address and a historical browsing page accessed according to the address, and webpage content data corresponding to the page are obtained; according to the method, the key value pair data set is constructed according to the webpage content data, the corresponding webpage content data can be rapidly obtained by constructing the key value pair data set, word segmentation processing is performed on the key value pair data set to obtain the multiple word segmentation data, and the searched result can be more comprehensive through the multiple word segmentation data; according to the word segmentation data and the query condition, the historical browsing page is subjected to query processing, the target query result is obtained, the target query result of the webpage can be automatically obtained, the data query difficulty is lowered, and therefore the query efficiency is improved.

The invention discloses a trademark retrieval system and method built based on a database. The system comprises a WEB server, a retrieval condition analysis unit, an image matching unit, a database server and a heterogeneous data memory. Mass image data in Internet is subjected to preprocessing, feature extraction and partitioned storage, the database is built and updated in real time, to-be-retrieved trademarks are subjected to preprocessing and feature extraction and matched with image data in the database updated in real time, the system and method are not restrict to a standard trademark library, and the retrieval range is comprehensive. Due to the fact that the to-be-retrieved trademarks and images in the database are subjected to preprocessing and feature extraction, matching featurepoints are more precise, the recognition error is small, and the retrieval accuracy and efficiency are greatly improved.

The invention discloses a retrieval system and method suitable for a combined trademark. The system comprises a WEB server, a first calculator, a text retrieval condition analysis unit, an image retrieval condition analysis unit, a text matching unit, an image matching unit and a heterogeneous data memory. The retrieval system is suitable for the combined trademark; the to-be-retrieved combined trademark is segmented into a text part and an image part; the text part and the image part are processed by utilizing the text retrieval condition analysis unit and the image retrieval condition analysis unit respectively; a to-be-retrieved text is subjected to keyword extraction; a to-be-retrieved image is subjected to preprocessing and feature extraction; the to-be-retrieved text and image are matched with massive text data and image data in the internet; and a trademark library is not limited to a standard trademark library, so that a retrieval range is wider. The to-be-retrieved text and image are subjected to the keyword extraction, the preprocessing and the feature extraction, so that matched feature points are more accurate, identification errors are small, and the retrieval accuracyand efficiency are greatly improved.

The invention provides an intellectual property operation platform system. The intellectual property operation platform system comprises a client and a system server, the system server comprises an intellectual property resource database; the client comprises a query module and a display module; the query module is used for obtaining keywords input by a client and sending the keywords to the system server; the system server is used for acquiring synonyms and synonyms of the keywords according to the keywords, retrieving in an intellectual property resource database according to the keywords and the synonyms and synonyms of the keywords to obtain retrieval results and sending the retrieval results to the client; and the client receives the retrieval result and displays the retrieval resultby the display module. Because customers are not professionals in the aspect of patents, input keywords of the customers are often deviated, and the problem that retrieval results are not comprehensive due to the deviation can be effectively solved.

The invention discloses a complete patent retrieval system. The complete patent retrieval system is characterized by comprising a first retrieval type storage unit; a first search unit; a complete search sample storage unit; a recall ratio calculation unit; a recall ratio comparison unit used for comparing a ratio obtained by the recall ratio calculation unit with a threshold value; an accuracy rate calculation unit used for performing spot check on literature from a retrieval result set to form a spot check literature set when the ratio obtained by the recall ratio calculation unit is greaterthan a threshold value, and comparing the number of literature related to the retrieval strategy in the spot check literature set with the number of literature in the spot check literature set to obtain a ratio; an accuracy comparison unit used for comparing the ratio obtained by the accuracy calculation unit with a threshold value; and an output unit used for outputting a retrieval result set when the ratio obtained by the accuracy calculation unit is greater than a threshold value. According to the method, a comprehensive and accurate retrieval result can be obtained, accurate patent analysis can be obtained, and decision errors are reduced.