Human-computer dialogue method and system for database index management based on large models

Abstract

The present invention discloses a human-machine dialogue method and system for database index management based on a large model; based on multiple matching statement segments, the similarity between pairwise matching statement segments is judged to obtain a set of non-similar statement segments and multiple sets of similar statement segments; based on the set of non-similar statement segments and multiple sets of similar statement segments, a first search engine including an index structure is constructed; during the interaction process, according to the constructed keyword association graph, through a deep learning network, answer data more suitable for the interaction statement is found; and the similarity of the answer data is detected to obtain a set of non-similar statement segments and multiple sets of similar statement segments; an index structure is constructed according to the set of non-similar statement segments and multiple sets of similar statement segments; the index structure can be used to construct a professional search engine for a vertical field of an industry, which is a subdivision and extension of the search engine.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular, to a human-computer dialogue method and system for database index management based on a large model. Background Art

[0002] At present, the combined application direction of artificial intelligence and data center has been very extensive. It can quickly develop intelligent agents for data analysis based on a local large model, and the front end can interact through virtual humans or other forms. The large model requires a large amount of data support, but the general search engine composed of the large model and the interactive front end has problems such as a large amount of information, inaccurate query, and insufficient depth. Summary of the Invention

[0003] The purpose of the present invention is to provide a human-computer dialogue method and system for database index management based on a large model to solve the above problems existing in the prior art.

[0004] In a first aspect, an embodiment of the present invention provides a human-computer dialogue method for database index management based on a large model, including:

[0005] Obtain an interaction statement; the interaction statement is a statement for the user to interact during human-computer interaction;

[0006] Input the interaction statement into a keyword neural network to detect keywords and obtain n first keywords;

[0007] Detect statement segments containing the first keywords through a matching search engine to obtain m search statement segments;

[0008] Calculate the number of first keywords in each search statement segment as the first keyword quantity; m*n first keyword quantities are obtained corresponding to the m search statement segments;

[0009] Perform pre-matching based on the n first keywords, the m search statement segments, and the corresponding first keyword quantities to obtain multiple matching statement segments; the matching statement segments represent the statement segments that match the interaction statement in the search engine;

[0010] Based on the multiple matching statement segments, judge the similarity between the pairwise matching statement segments to obtain a set of non-similar statement segments and multiple sets of similar statement segments; a set of similar statement segments contains multiple similar statement segments; the set of non-similar statement segments contains pairwise non-similar matching statement segments; the similar statement segments represent similar matching statement segments;

[0011] Construct a first search engine including an index structure based on the set of non-similar statement segments and the multiple sets of similar statement segments.

[0012] Optionally, performing pre-matching based on the n first keywords, m search statement segments, and the corresponding number of first keywords to obtain multiple matching statement segments, including:

[0013] Randomly obtain a first keyword as the third keyword;

[0014] Use the first keywords other than the third keyword as the second keywords to obtain a set of second keywords; the number of elements in the set of second keywords is n - 1;

[0015] Based on the n first keywords, m sets of second keywords, and the corresponding number of first keywords, construct a graph structure to obtain a keyword association graph;

[0016] m search statement segments correspondingly obtain m keyword association graphs;

[0017] Based on the m keyword association graphs and the m search statement segments, obtain multiple matching statement segments; the matching statement segments represent search statement segments associated with the interaction statement.

[0018] Optionally, the constructing a graph structure based on the n first keywords, m sets of second keywords, and the corresponding number of first keywords to obtain a keyword association graph includes:

[0019] Establish an association relationship between a first keyword and a second keyword in a corresponding set of second keywords to obtain a keyword pair; the keyword pair includes a first keyword and a second keyword;

[0020] 1 first keyword correspondingly obtains n - 1 keyword pairs; n first keywords correspondingly obtain n * (n - 1) keyword pairs;

[0021] Add the corresponding numbers of the two first keywords of the keyword pair to obtain a keyword pair value;

[0022] n * (n - 1) keyword pairs correspondingly obtain n * (n - 1) keyword pair values;

[0023] Detect two keyword pairs with the same first keyword and second keyword, and associate the two keyword pairs as the same keyword pair;

[0024] Average the keyword pair values corresponding to the same keyword pair to obtain an average keyword value;

[0025] n * (n - 1) keyword pairs correspondingly obtain n * (n - 1) / 2 average keyword values;

[0026] Based on the n first keywords and the corresponding average keyword values, obtain a keyword association graph.

[0027] Optionally, obtaining the keyword association graph based on the n first keywords and the corresponding average keyword values includes:

[0028] Construct a connected graph with the n first keywords as rows and columns;

[0029] Among them, the number of rows and columns of the connected graph is the same, and the first keywords corresponding to the rows and columns are the same;

[0030] Fill the average keyword value into the corresponding position in the connected graph according to the first keyword and the second keyword corresponding to the keyword pair, to obtain the keyword association graph.

[0031] Optionally, obtaining multiple matching statement segments based on the m keyword association graphs and the m search statement segments includes:

[0032] Input the keyword association graph into the first matching network to obtain the first matching degree feature;

[0033] Input the first matching degree feature into the first classification network to obtain the first matching degree value; m first matching degree values are obtained corresponding to the m keyword association graphs;

[0034] Take the keyword association graph with the first matching degree value greater than the matching threshold as the matching association graph;

[0035] Take the search statement segment corresponding to the matching association graph as the matching statement segment.

[0036] Optionally, the first matching network includes a first convolution kernel, a second convolution kernel, and a third convolution kernel;

[0037] The first convolution kernel is a 2*2 two-dimensional convolution kernel; the second convolution kernel is an n*2 two-dimensional convolution kernel; the third convolution kernel is a 2*n two-dimensional convolution kernel;

[0038] Convolve by moving the first convolution kernel in the row and column directions of the keyword association graph with a step size of 1 to obtain the first feature map;

[0039] Convolve by moving the second convolution kernel in the row direction of the keyword association graph with a step size of 1 to obtain the second feature vector;

[0040] Convolve by moving the third convolution kernel in the column direction of the keyword association graph with a step size of 1 to obtain the third feature vector;

[0041] Input the first feature map, the second feature vector, and the third feature vector into the first neural network to obtain the first matching degree feature.

[0042] Optionally, based on multiple matching statement segments, determining the similarity between pairwise matching statement segments to obtain a set of dissimilar statement segments and multiple sets of similar statement segments, including:

[0043] In the matching statement segments, detect the positions of the keywords to obtain the first key point sequence numbers; the first key point sequence numbers represent the sequence numbers of the keywords sorted according to the word order in the matching statement segments;

[0044] Sort the keywords corresponding to the matching statement segments according to the first key point sequence numbers to obtain a first detection sequence;

[0045] Input the pairwise first detection sequences into a discrimination network to judge the similarity and obtain multiple first sets of similar statement segments; the first sets of similar statement segments contain two similar matching statement segments;

[0046] Find the intersection of the multiple first sets of similar statement segments to obtain a set of similar statement segments;

[0047] Construct a set of dissimilar statement segments from the matching statement segments other than those in the set of similar statement segments.

[0048] Optionally, based on the set of dissimilar statement segments and multiple sets of similar statement segments, constructing a first search engine including an index structure, including:

[0049] The index structure includes a first index structure and a second index structure;

[0050] Take a matching statement segment in the set of similar statement segments as the first data to be answered;

[0051] Put the first data to be answered into the first index structure;

[0052] Take the matching statement segments in the set of dissimilar statement segments as the second data to be answered;

[0053] Put the second data to be answered into the second index structure;

[0054] Based on the first index structure and the second index structure, construct a first search engine including an index structure.

[0055] Optionally, based on the first index structure and the second index structure, constructing a first search engine including an index structure, including:

[0056] The first search engine outputs the first data to be answered in the first index structure;

[0057] If the first search engine obtains a second interaction statement; the second interaction statement represents a supplement to the interaction statement;

[0058] The first search engine will use the second index structure as a matching search engine to detect the second interaction statement and obtain a second set of similar statement segments; the second set of similar statement segments represents the similar matching statement segments corresponding to the second interaction statement.

[0059] In a second aspect, an embodiment of the present invention provides a human-computer dialogue system for database index management based on a large model, including:

[0060] An acquisition module, configured to acquire interaction statements; the interaction statements are statements for users to interact during human-computer interaction;

[0061] A keyword detection module, configured to input the interaction statements into a keyword neural network to detect keywords and obtain n first keywords;

[0062] A search module, configured to detect statement segments containing the first keywords through a matching search engine and obtain m search statement segments;

[0063] A quantity calculation module, configured to calculate the quantity of the first keywords in the search statement segments respectively as the first keyword quantities; m search statement segments correspond to obtaining m*n first keyword quantities;

[0064] A pre-matching module, based on the n first keywords, m search statement segments and the corresponding first keyword quantities, performs pre-matching to obtain multiple matching statement segments; the matching statement segments represent the statement segments in the matching search engine that answer the interaction statements;

[0065] A similarity judgment module, configured to judge the similarity between the matching statement segments pairwise based on the multiple matching statement segments, and obtain a set of non-similar statement segments and multiple sets of similar statement segments; a set of similar statement segments contains multiple similar statement segments; the set of non-similar statement segments contains pairwise non-similar matching statement segments; the similar statement segments represent similar matching statement segments;

[0066] A search engine construction module, configured to construct a first search engine including an index structure based on the set of non-similar statement segments and the multiple sets of similar statement segments.

[0067] Compared with the prior art, the embodiments of the present invention achieve the following beneficial effects:

[0068] In the present invention, during the interaction process, according to the constructed keyword association graph, through a deep learning network, answer data more suitable for the interaction statement is found. The similarity of the answer data is detected, and a set of dissimilar statement segments and multiple sets of similar statement segments are obtained. An index structure is constructed based on the set of dissimilar statement segments and the multiple sets of similar statement segments. The index structure can be used to construct a professional search engine for a vertical field of an industry, which is a refinement and extension of the search engine. A search engine service mode is obtained. Through this search engine service mode, search services can be provided for a certain type of information. The search engine service mode can search web page libraries and local data. The technical effect of obtaining the data required by the user according to the field orientation, processing it, and sending it to the user can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 It is a flowchart of a human-machine dialogue method for database index management based on a large model provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0070] The present invention will be described in detail below with reference to the accompanying drawings.

[0071] Embodiment 1

[0072] As Figure 1 shown, an embodiment of the present invention provides a human-machine dialogue method for database index management based on a large model, and the method includes:

[0073] S101: Obtain an interaction statement; the problem statement is a statement for the user to interact during human-machine interaction.

[0074] Wherein, in this embodiment, if the interaction statement is "What are the common loss functions in deep learning".

[0075] S102: Input the interaction statement into a keyword neural network to detect keywords and obtain n first keywords.

[0076] Wherein, any two of the first keywords are different keywords.

[0077] Wherein, the keyword neural network is a fully connected neural network (FullyConnectedNeturalNetwork, FCN).

[0078] Among them, the detected keywords are the keywords stored in the database. For example, if the interaction statement is "What are the common loss functions in deep learning", the database stores keywords such as "deep learning", "common", "loss function", and "include". Then, for the first keyword, n is 4, and the 4 first keywords are "deep learning", "common", "loss function", and "include" respectively. Use the keywords stored in the database as labeled data to train the keyword neural network.

[0079] Among them, n is a positive integer.

[0080] S103: Detect the statement segments containing the first keyword through a matching search engine to obtain m search statement segments.

[0081] Among them, m is a positive integer.

[0082] Among them, one search statement segment contains n first keywords.

[0083] Among them, the matching search engine can use the corresponding text in the web page that interacts with the interaction statement as a search statement segment. For example, the text between in html is used as a search statement segment.

[0084] S104: Calculate the number of the first keywords in the search statement segments respectively as the number of the first keywords; m search statement segments correspond to obtaining m * n numbers of the first keywords.

[0085] Among them, one keyword corresponds to one keyword number.

[0086] S105: Based on the n first keywords, m search statement segments and the corresponding numbers of the first keywords, perform pre - matching to obtain multiple matching statement segments; the matching statement segments represent the statement segments in the matching search engine that answer the interaction statement.

[0087] Among them, in this embodiment, the matching search engine includes a network search engine and a local search engine.

[0088] S106: Based on multiple matching statement segments, judge the similarity between pairwise matching statement segments to obtain a set of non - similar statement segments and multiple sets of similar statement segments; a set of similar statement segments contains multiple similar statement segments; the set of non - similar statement segments contains pairwise non - similar matching statement segments; the similar statement segments represent similar matching statement segments.

[0089] S107: Based on the set of non - similar statement segments and multiple sets of similar statement segments, construct a first search engine including an index structure.

[0090] Optionally, performing forward matching based on the n first keywords, m search statement segments, and the corresponding number of first keywords to obtain multiple matching statement segments, including:

[0091] Randomly obtain a first keyword as the third keyword;

[0092] Use the first keywords other than the third keyword as the second keywords to obtain a second keyword set; the number of elements in the second keyword set is n - 1.

[0093] Based on the n first keywords, m second keyword sets, and the corresponding number of first keywords, construct a graph structure to obtain a keyword association graph.

[0094] Among them, the keyword connection graph represents the degree of association relationship between the keywords in the search statement segment and the keywords in the interaction statement.

[0095] m search statement segments correspondingly obtain m keyword association graphs.

[0096] Based on the m keyword association graphs and the m search statement segments, obtain multiple matching statement segments; the matching statement segments represent the search statement segments associated with the interaction statement.

[0097] Optionally, the constructing a graph structure based on the n first keywords, m second keyword sets, and the corresponding number of first keywords to obtain a keyword association graph includes:

[0098] Establish an association relationship between a first keyword and a second keyword in a corresponding second keyword set to obtain a keyword pair; the keyword pair includes a first keyword and a second keyword.

[0099] Among them, in this embodiment, the association relationship in the keyword pair is established by storing the keyword pair as a set.

[0100] 1 first keyword correspondingly obtains n - 1 keyword pairs; n first keywords correspondingly obtain n * (n - 1) keyword pairs.

[0101] Add the corresponding number of first keywords of the keyword pair to obtain a keyword pair value.

[0102] n * (n - 1) keyword pairs correspondingly obtain n * (n - 1) keyword pair values.

[0103] Detect two keyword pairs with the same first keyword and second keyword, and associate the two keyword pairs as the same keyword pair.

[0104] Average the keyword pair values corresponding to the same key pair to obtain an average keyword value.

[0105] n*(n - 1) keyword pairs correspond to obtaining n*(n - 1) / 2 average keyword values.

[0106] Based on n first keywords and the corresponding average keyword values, a keyword association graph is obtained.

[0107] Optionally, the obtaining of the keyword association graph based on n first keywords and the corresponding average keyword values includes:

[0108] Taking n first keywords as rows and columns to construct a connected graph.

[0109] Among them, the number of rows and columns of the connected graph is the same, and the first keywords corresponding to the rows and columns are the same.

[0110] Filling the average keyword values into the corresponding positions in the connected graph according to the first keyword and the second keyword corresponding to the keyword pair, to obtain the keyword association graph.

[0111] Among them, by constructing the connected graph, it represents a graph structure constructed with n first keywords as nodes and average keyword values as edges.

[0112] Through the above method, a graph is used to establish the association relationship between n first keywords in the search statement segment.

[0113] Optionally, the obtaining of multiple matching statement segments based on m keyword association graphs and m search statement segments includes:

[0114] Inputting the keyword association graph into the first matching network to obtain the first matching degree feature.

[0115] Among them, the first matching network is a Convolutional Neural Networks (CNN).

[0116] Inputting the first matching degree feature into the first classification network to obtain the first matching degree value; m keyword association graphs correspond to obtaining m first matching degree values.

[0117] Among them, in this embodiment, the first classification network is a softmax function.

[0118] Among them, the range of the first matching degree value is from 0 to 1.

[0119] Taking the keyword association graph with the first matching degree value greater than the matching threshold as the matching association graph.

[0120] Among them, in this embodiment, the matching threshold is 0.85.

[0121] Taking the search statement segment corresponding to the matching association graph as the matching statement segment.

[0122] Optionally, the first matching network includes a first convolutional kernel, a second convolutional kernel, and a third convolutional kernel.

[0123] The first convolutional kernel is a 2*2 two-dimensional convolutional kernel; the second convolutional kernel is an n*2 two-dimensional convolutional kernel; the third convolutional kernel is a 2*n two-dimensional convolutional kernel.

[0124] In the row and column directions of the keyword association graph, the first convolutional kernel is moved with a step size of 1 for convolution to obtain a first feature map.

[0125] In the row direction of the keyword association graph, the second convolutional kernel is moved with a step size of 1 for convolution to obtain a second feature vector.

[0126] In the column direction of the keyword association graph, the third convolutional kernel is moved with a step size of 1 for convolution to obtain a third feature vector.

[0127] The first feature map, the second feature vector, and the third feature vector are input into a first neural network to obtain a first matching degree feature.

[0128] Among them, the first neural network is a fully connected neural network (Fully Connected Neural Network, abbreviated as FCNN), which can stretch the first feature map into a one-dimensional vector for convolution.

[0129] Through the above method, convolutional kernels with different structures are adopted, and three convolution methods are used. Relying on the symmetric characteristics of the keyword association graph, the association characteristics of the keyword association graph can be obtained more accurately.

[0130] m search statement segments correspond to obtaining m second keyword sets.

[0131] Optionally, the method for judging the similarity between pairwise matching statement segments based on multiple matching statement segments to obtain a set of dissimilar statement segments and multiple sets of similar statement segments includes:

[0132] In the matching statement segments, the positions of the first keywords are detected to obtain first key point numbers; the first key point numbers represent the numbers of the words sorted according to the word order in the matching statement segments.

[0133] Among them, the matching statement segment corresponding to the interactive statement "What are the common loss functions in deep learning?" is "The loss function in deep learning is used to measure the gap between the model prediction result and the actual label. Commonly used ones are: Mean Squared Error: applicable to regression tasks. Cross Entropy: often used in classification tasks. Binary Cross Entropy: specifically for binary classification problems.", because the keywords stored in the database include "Mean Squared Error", "Cross Entropy", and "Binary Cross Entropy". The first key point numbers are 1 corresponding to "Mean Squared Error", 2 corresponding to "Cross Entropy", and 3 corresponding to "Binary Cross Entropy".

[0134] Sort the first keywords corresponding to the matching statement segments according to the first key point numbers to obtain the first detection sequence.

[0135] Among them, in this embodiment, according to the first key point numbers, the first detection sequence is constructed as <"Mean Squared Error", "Cross Entropy", "Binary Cross Entropy">.

[0136] Input two first detection sequences into the discriminant network to judge the similarity and obtain the set of similar statement segments.

[0137] Among them, the discriminant network is the Discriminator.

[0138] Construct a set of non - similar statement segments from the matching statement segments other than those in the set of similar statement segments.

[0139] Among them, the set of non - similar statement segments represents a set containing multiple matching statement segments other than those in the set of similar statement segments.

[0140] Optionally, constructing a first search engine including an index structure based on the set of non - similar statement segments and multiple sets of similar statement segments includes:

[0141] The index structure includes a first index structure and a second index structure;

[0142] Take a matching statement segment in the set of similar statement segments as the first data to be answered;

[0143] Put the first data to be answered into the first index structure;

[0144] Take the matching statement segments in the set of non - similar statement segments as the second data to be answered;

[0145] Put the second data to be answered into the second index structure;

[0146] Construct a first search engine including an index structure based on the first index structure and the second index structure.

[0147] Optionally, building a first search engine including an index structure based on the first index structure and the second index structure includes:

[0148] The first search engine outputs the first data to be answered in the first index structure;

[0149] If the first search engine obtains a second interaction statement; the second interaction statement represents a supplement to the interaction statement;

[0150] The first search engine uses the second index structure as a matching search engine to detect the second interaction statement and obtains a second set of similar statement segments; the second set of similar statement segments represents the similar matching statement segments corresponding to the second interaction statement.

[0151] Among them, the detection method of the second interaction statement is the same as that of the interaction statement.

[0152] Embodiment 2

[0153] Based on the above human-machine dialogue method for database index management based on a large model, an embodiment of the present invention further provides a human-machine dialogue system for database index management based on a large model. The system includes an acquisition module, a keyword detection module, a search module, a quantity calculation module, a pre-matching module, a similarity judgment module, and a search engine construction module.

[0154] The acquisition module is used to acquire an interaction statement; the interaction statement is a statement for the user to interact during human-machine interaction;

[0155] The keyword detection module is used to input the interaction statement into a keyword neural network to detect keywords and obtain n first keywords;

[0156] The search module is used to detect statement segments containing the first keyword through a matching search engine and obtain m search statement segments;

[0157] The quantity calculation module is used to calculate the quantity of the first keyword in the search statement segment respectively as the first keyword quantity; m search statement segments correspond to obtaining m*n first keyword quantities;

[0158] The pre-matching module performs pre-matching based on the n first keywords, the m search statement segments, and the corresponding first keyword quantities to obtain multiple matching statement segments; the matching statement segments represent the statement segments in the matching search engine that answer the interaction statement;

[0159] A similarity judgment module, configured to judge the similarity between pairwise matching statement segments based on multiple matching statement segments, so as to obtain a set of dissimilar statement segments and multiple sets of similar statement segments; a set of similar statement segments contains multiple similar statement segments; the set of dissimilar statement segments contains pairwise dissimilar matching statement segments; the similar statement segments represent similar matching statement segments.

[0160] A search engine construction module, configured to construct a first search engine including an index structure based on the set of dissimilar statement segments and the multiple sets of similar statement segments.

[0161] Similarly, it should be understood that, in order to streamline the present disclosure and help understand one or more of the various inventive aspects, in the above description of the exemplary embodiments of the present invention, the various features of the present invention are sometimes grouped together into a single embodiment, figure, or description thereof. However, the disclosed method should not be construed as reflecting the intention that the claimed invention requires more features than those expressly recited in each claim. Rather, as reflected in the following claims, the inventive aspects lie in less than all the features of the single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into the detailed description, where each claim itself serves as a separate embodiment of the present invention.

Claims

1. A human-computer dialogue method for database index management based on a large model, characterized in that: include: Obtaining an interactive statement; the interactive statement is a statement for a user to interact during human-computer interaction; Input the interactive sentences into the keyword neural network, detect the keywords, and obtain n first keywords; By matching the search engine, the sentence segments containing the first keyword are detected to obtain m search sentence segments; Calculate the number of first keywords in the search statement segments respectively as the number of first keywords; m search statement segments correspond to obtaining m*n first keywords; Based on the n first keywords, the m search sentence segments and the corresponding number of first keywords, a pre-match is performed to obtain a plurality of matching sentence segments; The matching sentence segment represents a sentence segment that matches the answering interactive sentence in the search engine; Based on multiple matching sentence segments, the similarity between the matching sentence segments is determined to obtain a set of dissimilar sentence segments and a set of multiple similar sentence segments; A similar sentence segment set contains multiple similar sentence segments; The dissimilar sentence segment set includes two dissimilar matching sentence segments; The similar sentence segments represent similar matching sentence segments; Based on the dissimilar sentence segment set and the plurality of similar sentence segment sets, a first search engine including an index structure is constructed, including: The index structure includes a first index structure and a second index structure; Taking a matching sentence segment in the similar sentence segment set as the first data to be answered; Putting the first data to be answered into a first index structure; Using the matching sentence segments in the set of dissimilar sentence segments as the second data to be answered; Putting the second data to be answered into a second index structure; Based on the first index structure and the second index structure, a first search engine including an index structure is constructed, including: The first search engine outputs the first data to be answered in the first index structure; If the first search engine obtains a second interactive statement; the second interactive statement represents a supplement to the interactive statement; The first search engine will be used as a matching search engine in the second index structure to detect the second interactive sentence and obtain a second similar sentence segment set; the second similar sentence segment set represents similar matching sentence segments corresponding to the second interactive sentence.

2. According to the human-computer dialogue method for database index management based on a large model according to claim 1, it is characterized in that: The method performs a pre-match based on the n first keywords, the m search sentence segments and the corresponding number of first keywords to obtain a plurality of matching sentence segments, including: Randomly obtain a first keyword as the third keyword; The first keyword except the third keyword is used as the second keyword to obtain a second keyword set; the number of elements in the second keyword set is n-1; Based on the n first keywords, the m second keyword sets and the corresponding number of first keywords, a graph structure is constructed to obtain a keyword association graph; m search sentence segments correspond to m keyword association graphs; Based on the m keyword association graphs and the m search sentence segments, a plurality of matching sentence segments are obtained; the matching sentence segments represent the search sentence segments associated with the interactive sentence.

3. The human-computer dialogue method for database index management based on a large model according to claim 2 is characterized in that: The method of constructing a graph structure based on the n first keywords, the m second keyword sets and the corresponding number of first keywords to obtain a keyword association graph includes: Establishing an association relationship between the first keyword and a second keyword in the corresponding second keyword set to obtain a keyword pair; the keyword pair includes a first keyword and a second keyword; One first keyword corresponds to n-1 keyword pairs; n first keywords correspond to n*(n-1) keyword pairs; Add the two first keyword quantities corresponding to the keyword pair to obtain the keyword pair value; n*(n-1) keyword pairs correspond to n*(n-1) keyword pair values; Detecting two keyword pairs in which the first keyword and the second keyword are identical, and associating the two keyword pairs as identical keyword pairs; Average the keyword pair values ​​corresponding to the same keyword pair to obtain the average keyword value; For n*(n-1) keyword pairs, n*(n-1) / 2 average keyword values ​​are obtained. Based on the n first keywords and the corresponding average keyword values, a keyword association graph is obtained.

4. The human-computer dialogue method for database index management based on a large model according to claim 2 is characterized in that: The step of obtaining a keyword association graph based on the n first keywords and the corresponding average keyword values ​​includes: Use the n first keywords as rows and columns to construct a connected graph; The number of rows and columns of the connected graph is the same, and the first keywords corresponding to the rows and columns are the same; According to the first keyword and the second keyword corresponding to the keyword pair, the average keyword value is filled into the corresponding position in the connected graph to obtain a keyword association graph.

5. The human-computer dialogue method for database index management based on a large model according to claim 2 is characterized in that: The method of obtaining multiple matching sentence segments based on the m keyword association graphs and the m search sentence segments includes: Inputting the keyword association graph into a first matching network to obtain a first matching degree feature; Inputting the first matching degree feature into the first classification network to obtain a first matching degree value; and obtaining m first matching degree values ​​corresponding to the m keyword association graphs; The keyword association graph whose first matching degree value is greater than the matching threshold is used as the matching association graph; The search sentence segment corresponding to the matching association graph is taken as the matching sentence segment.

6. The human-computer dialogue method for database index management based on a large model according to claim 5 is characterized in that: The first matching network includes a first convolution kernel, a second convolution kernel and a third convolution kernel; The first convolution kernel is a 2*2 two-dimensional convolution kernel; the second convolution kernel is an n*2 two-dimensional convolution kernel; the third convolution kernel is a 2*n two-dimensional convolution kernel; In the direction of rows and columns of the keyword association graph, the first convolution kernel is moved with a step size of 1 for convolution to obtain the first feature graph; In the row direction of the keyword association graph, the second convolution kernel is moved with a step size of 1 to perform convolution to obtain a second eigenvector; In the column direction of the keyword association graph, the third convolution kernel is moved with a step size of 1 to perform convolution to obtain a third eigenvector; The first feature map, the second feature vector and the third feature vector are input into a first neural network to obtain a first matching degree feature.

7. The human-computer dialogue method for database index management based on a large model according to claim 1 is characterized in that: The method of judging the similarity between the matching sentence segments based on the multiple matching sentence segments to obtain a set of dissimilar sentence segments and a set of similar sentence segments includes: In the matching sentence segment, the position of the keyword is detected to obtain a first key point sequence number; the first key point sequence number represents the sequence number of the keyword in the matching sentence segment sorted according to the word order; According to the first key point sequence number, the keywords corresponding to the matching sentence segments are sorted to obtain a first detection sequence; Inputting the first detection sequences in pairs into the discriminant network to judge the similarity, and obtaining a plurality of first similar sentence segment sets; the first similar sentence segment sets include two similar matching sentence segments; Obtaining an intersection of multiple first similar sentence segment sets to obtain a similar sentence segment set; The matching sentence segments other than the matching sentence segments in the similar sentence segment set are used to construct a dissimilar sentence segment set.

8. A human-computer dialogue system for database index management based on a large model, characterized in that: include: The acquisition module is used to obtain the interactive statements; The interactive sentences are sentences used by users to interact during human-computer interaction; A keyword detection module, used for inputting the interactive sentence into the keyword neural network, detecting the keyword, and obtaining n first keywords; A search module, used to detect a sentence segment containing the first keyword by matching a search engine, and obtain m search sentence segments; A quantity calculation module, used to calculate the quantity of the first keywords in the search sentence segments as the first keyword quantity; m search statement segments correspond to obtaining m*n first keywords; A pre-matching module performs pre-matching based on the n first keywords, the m search sentence segments and the corresponding number of first keywords to obtain a plurality of matching sentence segments; The matching sentence segment represents a sentence segment that matches the answering interactive sentence in the search engine; A similarity judgment module is used to judge the similarity between the matching sentence segments based on the multiple matching sentence segments, and obtain a set of dissimilar sentence segments and a set of multiple similar sentence segments; A similar sentence segment set contains multiple similar sentence segments; The dissimilar sentence segment set includes two dissimilar matching sentence segments; The similar sentence segments represent similar matching sentence segments; A search engine construction module is used to construct a first search engine including an index structure based on the non-similar sentence segment set and multiple similar sentence segment sets, including: The index structure includes a first index structure and a second index structure; Taking a matching sentence segment in the similar sentence segment set as the first data to be answered; Putting the first data to be answered into a first index structure; Using the matching sentence segments in the set of dissimilar sentence segments as the second data to be answered; Putting the second data to be answered into a second index structure; Based on the first index structure and the second index structure, constructing a first search engine including the index structure includes: The first search engine outputs the first data to be answered in the first index structure; If the first search engine obtains a second interactive statement; the second interactive statement represents a supplement to the interactive statement; The first search engine will be used as a matching search engine in the second index structure to detect the second interactive sentence and obtain a second similar sentence segment set; the second similar sentence segment set represents similar matching sentence segments corresponding to the second interactive sentence.

Images