A text matching method and device
A matching method and text technology, applied in the computer field, can solve problems such as inaccurate matching results and inconsistent positions.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
no. 1 example
[0080] see Figure 1A , is a schematic flowchart of a text matching method provided in this embodiment. The text matching method includes the following steps:
[0081] S101: Obtain the first text and the second text to be matched.
[0082] For ease of distinction, in this embodiment, the two texts to be matched are respectively defined as the first text and the second text. For example, in a search scenario, the text entered by the user can be used as the first text, and each text in the corpus can be used as the second text; or, each text in the corpus can be used as the first text, and the text entered by the user text as the second text. For another example, in the FAQ system, the questions raised by the user are used as the first text, and each question in the FAQ library is used as the second text respectively; or, each text in the FAQ library is used as the first text respectively, and the user Enter the question as the second text.
[0083] In order to facilitate th...
no. 2 example
[0096] see Figure 1B , is a schematic flowchart of a text matching method provided in this embodiment. The text matching method is mainly for the specific introduction of S104 in the first embodiment, and specifically includes the following steps:
[0097] S1041: Draw a feature matrix, the number of rows of the feature matrix is the total number of the first participle, and the number of columns of the feature matrix is the total number of the second participle.
[0098] Determine the total number Row of each first participle in the first text, and determine the total number Col of each second participle in the second text; afterward, draw the feature matrix M, the number of rows and the number of columns of the feature matrix M are respectively for Row and Col.
[0099] Such as figure 2 The schematic diagram of the feature matrix shown in the above example continues the above example, taking the number of word segments of the text S1 as the number of rows, and taking t...
no. 3 example
[0121] see Figure 4 , is a schematic flowchart of a text matching method provided in this embodiment. The text matching method is mainly specifically introduced for S1045 in the first embodiment, and specifically includes the following steps:
[0122] S401: If the size of the feature value is positively correlated with the degree of similarity, then when the feature value is greater than a first feature threshold, keep the feature value.
[0123] In this embodiment, for each eigenvalue in the eigenmatrix M, the eigenvalues corresponding to each group of words with higher similarity are retained. Specifically, when the value range of the feature value is [a1, b1], if the size of the feature value is positively correlated with the degree of similarity, then the first feature threshold may be δ1=(b1-a1 )*θ1%, where θ1% is a value greater than 50%, such as θ1=85%, and the eigenvalues greater than δ1 are reserved in the characteristic matrix M.
[0124] For example, for i...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More