Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Method and system for intelligently understanding user query intention

A technology for querying intentions and users, applied in the field of intelligently understanding user querying intentions, can solve problems such as incorrect acquisition of user querying intentions, failure to make better use of useful information, and different levels of thinking, so as to reduce the cumbersomeness of system operations, Improve word segmentation efficiency and improve processing speed

Inactive Publication Date: 2019-10-08
鼎复数据科技(北京)有限公司
View PDF8 Cites 12 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] (1) Most of the existing query search methods are queries based on short keywords or specific format templates, and the input length of the query is very limited. In the case of inputting a long text, it will be truncated and ignored in most cases. , so that the user's query intent cannot be obtained correctly;
[0005] (2) In the query algorithm for inputting complete sentences, there is no good use of the useful information brought by the key entities and syntactic structures in the sentences
Since a large amount of text is unstructured or semi-structured data, and the level of thinking of the people who write the documents is different, people need to understand and check all the content during the review process, but the content that actually needs to be focused on is actually not the same. There are not many or people in different departments pay attention to different content. For example, in financial statements, there are a lot of unstructured data, but they often pay more attention to each indicator and corresponding value without reading all the text content, resulting in a serious waste of manpower; Furthermore, it may be necessary to convert unstructured or semi-structured data into structured data, or analyze information pairs in unstructured or semi-structured data to obtain matching indicators and corresponding values

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Method and system for intelligently understanding user query intention
  • Method and system for intelligently understanding user query intention
  • Method and system for intelligently understanding user query intention

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0109] Taking the input query statement "China Merchants Bank's net profit and operating income" as an example, by understanding the query statement, we hope to obtain the user's real query intention:

[0110] The first step is to combine the dictionary and obtain word segmentation results through the forward maximum matching method:

[0111] China Merchants / Bank / Net Profit / Operation / Revenue / ;

[0112] The second step is to perform part-of-speech tagging on the part-of-speech results through the hidden Markov model obtained through training. The part-of-speech tagging results are:

[0113] China Merchants_v Bank_n Net Profit_n Business_n Revenue_v;

[0114] In the third step, the conditional random field model obtained through training is used for marking and named entity recognition, and the result is:

[0115] Marking: Merchants bank net profit business income

[0116] Entity identification: China Merchants Bank

[0117] In the fourth step, the grammatical rule...

Embodiment 2

[0122] Enter the query statement "As of March 31, 2016, the company's total liabilities were 1.036 billion, mainly composed of: short-term loans (including long-term loans due this year) 960 million, long-term loans 55 million, and accounts payable 7.07 million RMB 510,000, tax payable is RMB 510,000. The current loan scale is RMB 1.015 billion, and short-term loans account for 93% of the total liabilities, indicating that the company has a relatively large debt repayment pressure in the short term. Combined with the company’s existing monetary funds of RMB 762 million From the looks of it, the financial risk is not big." As an example, by understanding the query statement, we hope to obtain the user's real query intention:

[0123] The first step is to combine the dictionary and obtain word segmentation results through the forward maximum matching method:

[0124] Short-term / loan / (including / long-term / loan / of / this year / due / ) / 960 million / , / long-term / loan / 55 million / yuan / ...

Embodiment 3

[0139] The input query sentence is the same as that in Embodiment 2, and the method for understanding the user's query intention is the same as that in Embodiment 2, the only difference being that word segmentation results are obtained through the conditional random field model. The effect of conditional random field word segmentation model is shown in Table 9.

[0140] Table 9 Conditional Random Field Word Segmentation Model Effect

[0141] data set time Accuracy recall rate F value pku_test (510KB) 1.676s 0.931 0.919 0.925 msr_test (560KB) 1.928s 0.859 0.894 0.876

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a method and system for intelligently understanding a user query intention, and the method comprises the steps: inputting a query statement, and carrying out the word segmentation processing through combining with a dictionary; performing part-of-speech tagging on the word segmentation result; performing named entity recognition on the words after the part-of-speech tagging; and carrying out grammatical analysis through a named entity identification result and a set grammatical rule to obtain a user query intention. According to the method, the input query statements are analyzed layer by layer according to the wording characteristics in the loan auditing industry, the query intention of the user is deeply understood, and the query efficiency is improved on the premise that the accuracy is ensured.

Description

technical field [0001] The invention relates to natural language processing technology, in particular to a method and system for intelligently understanding user query intentions. Background technique [0002] The understanding and processing of user query intent aims to model, analyze and process user input queries. Understanding the intent of user queries will help improve the quality of information retrieval and user experience. The existing general search is characterized by crawling all valuable information on the Internet / database and building an index at the same time, with keyword matching as the basic retrieval method. In the traditional general search engine, because it needs to apply to a wide range of requirements, its intelligence is often not high; because improving its intelligence will greatly reduce the efficiency of search and make the search engine overwhelmed. Therefore, general-purpose search engines often have many defects when searching for informati...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F16/9535G06F17/27
CPCG06F40/295G06F40/30G06F16/9535
Inventor 杨云飞李超吴雪军
Owner 鼎复数据科技(北京)有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products