阅读说明：本技术 工程施工安全管理文档文本智能检索方法及装置 (Intelligent search method and device for engineering construction safety management document text ) 是由 李明超 田丹 沈扬 韩帅 任秋兵 于 2020-12-24 设计创作，主要内容包括：本公开提供了一种工程施工安全管理文档文本智能检索模型的训练方法及装置、文本智能检索方法及装置、以及计算机系统和计算机可读存储介质；其训练方法,包括：采集施工安全样本数据；划分为训练集和测试集,并进行预处理；将训练集中安全问题文本词向量集合和安全规范文本词向量集合分别输入至孪生神经网络两个子网络,得到安全问题特征向量集合和安全规范文本特征向量集合；确定安全问题特征向量和安全规范文本特征向量间的关联值,构建损失函数并计算损失值；将测试集中的安全问题文本词向量和安全规范文本词向量分别输入至两个子网络,对孪生神经网络参数进行训练,得到工程施工安全管理文档文本智能检索模型。(The present disclosure provides a training method and apparatus for an intelligent search model of engineering construction safety management document texts, an intelligent search method and apparatus for texts, a computer system and a computer-readable storage medium; the training method comprises the following steps: collecting construction safety sample data; dividing the training set and the test set, and preprocessing; respectively inputting the safety problem text word vector set and the safety standard text word vector set in the training set into two sub-networks of a twin neural network to obtain a safety problem characteristic vector set and a safety standard text characteristic vector set; determining a correlation value between the safety problem feature vector and the safety standard text feature vector, constructing a loss function and calculating a loss value; and respectively inputting the safety problem text word vectors and the safety standard text word vectors which are concentrated in the test into the two sub-networks, and training the parameters of the twin neural network to obtain an intelligent text retrieval model of the engineering construction safety management document.)

1. A training method for an intelligent text retrieval model of engineering construction safety management documents comprises the following steps:

collecting sample data; wherein the sample data is characterized by a plurality of text pairs; the text pairs include: a security problem text and a security specification text which are mutually related;

dividing the sample data into a training set and a testing set, and preprocessing the training set and the testing set to obtain a safety problem text word vector set and a safety standard text word vector set in the training set and the testing set;

respectively inputting the safety problem text word vector set and the safety standard text word vector set in the training set into a first sub-network and a second sub-network of a twin neural network to obtain a safety problem feature vector set and a safety standard text feature vector set, wherein the safety problem feature vector set and the safety standard text feature vector set are used for representing the result of text depth analysis; wherein the first sub-network and the second sub-network of the twin neural network are both generated based on a Bi-LSTM model;

determining a correlation value between a safety problem characteristic vector and a safety standard text characteristic vector according to the safety problem characteristic vector set and the safety standard text characteristic vector set, constructing a loss function and calculating a loss value; and

and respectively inputting the safety problem text word vector set and the safety standard text word vector set in the test set into the first sub-network and the second sub-network of the twin neural network, and performing optimization training on the first sub-network and the second sub-network of the twin neural network to obtain the intelligent text retrieval model of the engineering construction safety management document.

2. The training method according to claim 1, wherein the dividing the sample data into a training set and a test set, and preprocessing the training set and the test set to obtain a safety problem text word vector set and a safety specification text word vector set in the training set and the test set comprises:

dividing the sample data into a training set and a test set according to a set proportion; wherein the training set and the test set include at least one of the sample data;

performing Jieba word segmentation on each safety problem text and each safety standard text in the training set based on a Jieba word segmentation library to obtain a plurality of text words in each safety problem text and each safety standard text;

analyzing the main body structure of each safety problem text and each safety standard text in the training set by utilizing a semantic structure model, defining the property of each text word in each safety problem text and each safety standard text, and labeling the text words;

rejecting each safety problem text and text symbol in the safety standard text in the training set and the test set; and

and performing text word vector calculation on each safety problem text and each safety standard text in the training set and the test set by using a skip-gram model to obtain a safety problem text word vector set and a safety standard text word vector set in the training set and the test set.

3. The training method according to claim 2, wherein the performing text word vector calculation on each of the safety issue texts and the safety standard texts in the training set and the test set by using a skip-gram model to obtain a safety issue text word vector set and a safety standard text word vector set in the training set and the test set further comprises:

determining the weight of each text word vector in the safety problem text word vector set and the safety standard text word vector set in the training set and the test set by utilizing an Attention mechanism, and integrating the weight into the corresponding text word vector to obtain the safety problem text word vector and the safety standard text word vector considering the weight.

4. The training method according to claim 1, wherein the set of safety problem text word vectors and the set of safety standard text word vectors in the training set are respectively input into a first subnetwork and a second subnetwork of a twin neural network, so as to obtain a set of safety problem feature vectors and a set of safety standard text feature vectors, and the set of safety problem feature vectors and the set of safety standard text feature vectors are used for characterizing a result of text depth analysis; wherein the first sub-network and the second sub-network of the twin neural network are each generated based on a Bi-LSTM model, the repeating comprising:

the safety problem text word vector sequentially passes through an input gate, a forgetting gate and an output gate of the Bi-LSTM model corresponding to the first sub-network in the twin neural network; the safety standard text word vector sequentially passes through an input gate, a forgetting gate and an output gate of the Bi-LSTM model corresponding to the second subnetwork in the twin neural network; and

traversing the safety problem text word vector along a first direction and a second direction by the Bi-LSTM model corresponding to the first sub-network to obtain a safety problem feature vector; traversing the safety standard text word vector along a first direction and a second direction by the Bi-LSTM model corresponding to the second sub-network to obtain a safety standard text feature vector; wherein the first direction represents a forward word order direction and the second direction represents a reverse word order direction;

and obtaining a security problem feature vector set and a security specification text feature vector set.

5. The training method of claim 1, wherein the constructing a loss function from the set of security problem feature vectors and the set of security specification text feature vectors comprises:

calculating the correlation value between the safety problem feature vector and the safety specification text feature vector output by the first sub-network and the second sub-network by using the Manhattan distance, wherein the formula is as follows:

Dis(X₁，X₂)＝|x₁-x₂|+|y₁-y₂|

in the formula, X₁、X₂Respectively representing a safety problem feature vector and a safety standard text feature vector; x is the number of₁、y₁Two-dimensional space coordinates of the safety problem feature vector; x is the number of₂、y₂Two-dimensional space coordinates of the text feature vector are specified for safety; dis (X)₁，X₂) The Manhattan distance between the safety problem feature vector and the safety standard text feature vector is the correlation value between the safety problem feature vector and the safety standard text feature vector; and

calculating a Loss value by using a contrast Loss function, wherein the formula is as follows:

in the formula: n is the number of the safety problem characteristic vectors and the safety standard text characteristic vectors in the training set; when the sample dissimilarity threshold m is met, the safety problem feature vector is dissimilar to the safety specification text feature vector, and Y is 0; when the sample dissimilarity threshold m is not satisfied, the security problem feature vector is similar to the security specification text feature vector, and Y is 1.

6. An intelligent search method for engineering construction safety management document texts comprises the following steps:

acquiring an intelligent search model of an engineering construction safety management document text; and

extracting a safety standard text matched with the input safety problem text through an engineering construction safety management document text intelligent retrieval model, and outputting a correlation value between the safety problem text input by the number of pieces and the safety standard text;

the intelligent search model of the engineering construction safety management document text is obtained by training through the training method of any one of claims 1 to 5.

7. A training device for an intelligent text retrieval model of engineering construction safety management documents comprises:

the first acquisition module is used for acquiring sample data; wherein the sample data is characterized by a plurality of text pairs; the text pairs include: a security problem text and a security specification text which are mutually related;

the preprocessing module is used for dividing the sample data into a training set and a testing set, and preprocessing the training set and the testing set to obtain a safety problem text word vector set and a safety standard text word vector set in the training set and the testing set;

a data set construction module, configured to input the safety problem text word vector set and the safety standard text word vector set in the training set into a first subnetwork and a second subnetwork of the twin neural network, respectively, to obtain a safety problem feature vector set and a safety standard text feature vector set, where the safety problem feature vector set and the safety standard text feature vector set are used to represent a result of text depth analysis; wherein the first sub-network and the second sub-network of the twin neural network are both generated based on a Bi-LSTM model;

the loss function building module is used for determining an associated value between a safety problem feature vector and a safety standard text feature vector according to the safety problem feature vector set and the safety standard text feature vector set, building a loss function and calculating a loss value; and

and the training module is used for respectively inputting the safety problem text word vector set and the safety standard text word vector set in the test set into the first sub-network and the second sub-network of the twin neural network, and performing optimization training on the first sub-network and the second sub-network of the twin neural network to obtain the intelligent text retrieval model of the engineering construction safety management document.

8. An intelligent text retrieval device for engineering construction safety management documents comprises:

the second acquisition module is used for acquiring an intelligent search model of the engineering construction safety management document text; and

the retrieval module is used for extracting a safety standard text matched with the input safety problem text through an engineering construction safety management document text intelligent retrieval model and outputting a correlation value between the safety problem text input by the number of pieces and the safety standard text;

the intelligent search model of the engineering construction safety management document text is obtained by training through the training method of any one of claims 1 to 5.

9. A computer system, comprising:

one or more processors;

a memory for storing one or more programs,

wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of any of claims 1-5.

10. A computer readable storage medium having stored thereon executable instructions which, when executed by a processor, cause the processor to carry out the method of any one of claims 1 to 5.

Technical Field

The disclosure relates to the field of construction safety management of large-scale infrastructure projects such as water conservancy projects, building projects and traffic transportation projects, in particular to a training method and device of a semantic-based intelligent search model of an engineering construction safety management document text, a text search method and device, a computer system and a computer-readable storage medium.

Background

The construction of large-scale foundation construction engineering is a complex system engineering, and has the characteristics of long engineering construction time, huge engineering quantity, huge resource consumption, complex construction process, high construction strength, continuous and rapid construction process, high construction quality control difficulty, high technical requirements and the like, so that the construction site safety management difficulty is increased more and more. Aiming at large-scale engineering construction projects, the perfect safety management system can not only effectively ensure the health of constructors, but also improve the engineering construction efficiency. With the increasing requirements of engineering construction on safety management, more and more safety management measures are proposed and applied to the actual construction process. In the implementation process of a large number of safety management measures, massive construction safety management texts are generated, the safety management conditions of a construction site are recorded, and a lot of safety management experiences are included. Meanwhile, in order to standardize the safety management process of a construction site, a large number of safety management specifications and standards related to the industry are issued by related national departments and the industry, and the safety management system comprises a plurality of new ideas and new methods for safety management, so that the safety management process of the site can be effectively and accurately guided.

In the practical application process, the utilization rate of the construction safety management document text is low, and a large amount of hidden key information is not mined. In order to improve the utilization efficiency of the construction safety management document text, a manager establishes a text information base for the system management of text data. With the development of artificial intelligence technology, text management begins to break away from the original artificial processing mode, and the text content is deeply mined by using an intelligent method and technology, so that the acquisition efficiency of construction safety management document knowledge is improved, and the real-time performance of safety management work is ensured.

In the construction process of large-scale infrastructure engineering, a plurality of sudden and high-risk potential safety hazards and safety management problems exist, and corresponding solutions need to be found and given in time. A large number of safety management measures are given by the existing construction safety management standard, and the site safety management task can be effectively guided. Because the safety management standards are mostly presented in the form of unstructured text data and are huge in number, aiming at specific safety management problems, the manual searching method is time-consuming and labor-consuming, and the timeliness of safety management of a construction site cannot be met. Meanwhile, the safety management problems of the construction site are diversified, comprehensive and multidirectional consideration and analysis are needed, and the difficulty of searching the safety management specifications is increased. Although the existing engineering construction text management has made great progress, in the intelligent processing process of the construction site safety management text, intensive research and study are needed.

Disclosure of Invention

Technical problem to be solved

The present disclosure provides a training method and apparatus for an intelligent search model of engineering construction safety management document text, a text search method and apparatus, a computer system, and a computer-readable storage medium, so as to solve the above-mentioned technical problems.

(II) technical scheme

According to one aspect of the disclosure, a training method for an intelligent text retrieval model of an engineering construction safety management document is provided, which includes:

collecting sample data; wherein the sample data is characterized by a plurality of text pairs; the text pairs include: a security problem text and a security specification text which are mutually related;

dividing the sample data into a training set and a testing set, and preprocessing the training set and the testing set to obtain a safety problem text word vector set and a safety standard text word vector set in the training set and the testing set;

respectively inputting the safety problem text word vector set and the safety standard text word vector set in a training set into a first sub-network and a second sub-network of the twin neural network to obtain a safety problem feature vector set and a safety standard text feature vector set, wherein the safety problem feature vector set and the safety standard text feature vector set are used for representing the result of text depth analysis; wherein the first sub-network and the second sub-network of the twin neural network are both generated based on a Bi-LSTM model;

determining a correlation value between a safety problem characteristic vector and a safety standard text characteristic vector according to the safety problem characteristic vector set and the safety standard text characteristic vector set, constructing a loss function and calculating a loss value; and

and respectively inputting the safety problem text word vector set and the safety standard text word vector set in the test set into a first sub-network and a second sub-network of the twin neural network, and performing optimization training on the first sub-network and the second sub-network of the twin neural network to obtain the intelligent text retrieval model of the engineering construction safety management document.

In some embodiments of the present disclosure, the dividing the sample data into a training set and a test set, and preprocessing the training set and the test set to obtain a security problem text word vector set and a security standard text word vector set in the training set and the test set includes:

dividing the sample data into a training set and a test set according to a set proportion; wherein the training set and the test set comprise at least one sample data;

performing Jieba word segmentation on each safety problem text and each safety standard text in the training set based on a Jieba word segmentation library to obtain a plurality of text words in each safety problem text and each safety standard text;

analyzing the main body structure of each safety problem text and each safety standard text in the training set by utilizing a semantic structure model, defining the property of each text word in each safety problem text and each safety standard text, and labeling the text words;

rejecting each safety problem text and text symbol in the safety standard text in the training set and the test set; and

and performing text word vector calculation on each safety problem text and each safety standard text in the training set and the test set by using a skip-gram model to obtain a safety problem text word vector set and a safety standard text word vector set in the training set and the test set.

In some embodiments of the disclosure, the performing, by using a skip-gram model, text word vector calculation on each of the safety problem texts and the safety standard texts in the training set and the test set to obtain a safety problem text word vector set and a safety standard text word vector set in the training set and the test set further includes:

determining the weight of each text word vector in the safety problem text word vector set and the safety standard text word vector set in the training set and the test set by utilizing an Attention mechanism, and integrating the weight into the corresponding text word vector to obtain the safety problem text word vector and the safety standard text word vector considering the weight.

In some embodiments of the present disclosure, the inputting the set of security problem text word vectors and the set of security specification text word vectors in the training set into a first subnetwork and a second subnetwork of the twin neural network, respectively, obtains a set of security problem feature vectors and a set of security specification text feature vectors, and the set of security problem feature vectors and the set of security specification text feature vectors are used for characterizing a result of text depth analysis; wherein the first sub-network and the second sub-network of the twin neural network are each generated based on a Bi-LSTM model, the repeating comprising:

the safety problem text word vector sequentially passes through an input gate, a forgetting gate and an output gate of the Bi-LSTM model corresponding to the first sub-network in the twin neural network; the safety standard text word vector sequentially passes through an input gate, a forgetting gate and an output gate of the Bi-LSTM model corresponding to the second subnetwork in the twin neural network; and

traversing the safety problem text word vector along a first direction and a second direction by the Bi-LSTM model corresponding to the first sub-network to obtain a safety problem feature vector; traversing the safety standard text word vector along a first direction and a second direction by the Bi-LSTM model corresponding to the second sub-network to obtain a safety standard text feature vector; wherein the first direction represents a forward word order direction and the second direction represents a reverse word order direction;

and obtaining a security problem feature vector set and a security specification text feature vector set.

In some embodiments of the disclosure, constructing the loss function according to the set of security problem feature vectors and the set of security specification text feature vectors includes:

calculating the correlation value between the safety problem feature vector and the safety specification text feature vector output by the first sub-network and the second sub-network by using the Manhattan distance, wherein the formula is as follows:

Dis(X₁，X₂)＝|x₁-x₂|+|y₁-y₂|

in the formula, X₁、X₂Respectively representing a safety problem feature vector and a safety standard text feature vector; x is the number of₁、y₁Two-dimensional space coordinates of the safety problem feature vector; x is the number of₂、y₂Two-dimensional space coordinates of the text feature vector are specified for safety; dis (X)₁，X₂) The Manhattan distance between the safety problem feature vector and the safety standard text feature vector is the correlation value between the safety problem feature vector and the safety standard text feature vector; and

calculating a Loss value by using a contrast Loss function, wherein the formula is as follows:

in the formula: n is the number of the safety problem characteristic vectors and the safety standard text characteristic vectors in the training set; when the sample dissimilarity threshold m is met, the safety problem feature vector is dissimilar to the safety specification text feature vector, and Y is 0; when the sample dissimilarity threshold m is not satisfied, the security problem feature vector is similar to the security specification text feature vector, and Y is 1.

According to one aspect of the disclosure, an intelligent text retrieval method for engineering construction safety management documents is provided, which includes:

acquiring an intelligent search model of an engineering construction safety management document text; and

extracting a safety standard text matched with the input safety problem text through an engineering construction safety management document text intelligent retrieval model, and outputting a correlation value between the safety problem text input by the number of pieces and the safety standard text;

the intelligent search model of the engineering construction safety management document text is obtained by training through the training method.

According to one aspect of the disclosure, a training device for an intelligent text retrieval model of an engineering construction safety management document is provided, which includes:

the first acquisition module is used for acquiring sample data; wherein the sample data is characterized by a plurality of text pairs; the text pairs include: a security problem text and a security specification text which are mutually related;

the preprocessing module is used for dividing the sample data into a training set and a testing set, and preprocessing the training set and the testing set to obtain a safety problem text word vector set and a safety standard text word vector set in the training set and the testing set;

the data set construction module is used for respectively inputting the safety problem text word vector set and the safety standard text word vector set in a training set into a first sub-network and a second sub-network of the twin neural network to obtain a safety problem feature vector set and a safety standard text feature vector set, and the safety problem feature vector set and the safety standard text feature vector set are used for representing the result of text depth analysis; wherein the first sub-network and the second sub-network of the twin neural network are both generated based on a Bi-LSTM model;

the loss function building module is used for determining an associated value between a safety problem feature vector and a safety standard text feature vector according to the safety problem feature vector set and the safety standard text feature vector set, building a loss function and calculating a loss value; and

and the training module is used for respectively inputting the safety problem text word vector set and the safety standard text word vector set in the test set into a first sub-network and a second sub-network of the twin neural network, and performing optimization training on the first sub-network and the second sub-network of the twin neural network to obtain the intelligent text retrieval model of the engineering construction safety management document.

According to one aspect of the present disclosure, there is provided an intelligent text retrieval device for engineering construction safety management documents, comprising:

the second acquisition module is used for acquiring an intelligent search model of the engineering construction safety management document text; and

the retrieval module is used for extracting a safety standard text matched with the input safety problem text through an engineering construction safety management document text intelligent retrieval model and outputting a correlation value between the safety problem text input by the number of pieces and the safety standard text;

the intelligent search model of the engineering construction safety management document text is obtained by training through the training method.

According to an aspect of the present disclosure, there is provided a computer system including:

one or more processors;

a memory for storing one or more programs,

wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method described above.

According to an aspect of the present disclosure, there is provided a computer-readable storage medium having stored thereon executable instructions that, when executed by a processor, cause the processor to implement the above-described method.

(III) advantageous effects

According to the technical scheme, the intelligent text retrieval method for the engineering construction safety management document based on the deep semantic analysis has at least one or part of the following beneficial effects:

(1) based on the text intelligent retrieval model provided by the disclosure, safety problems in engineering construction can be found quickly, safety management experience and problem solutions can be retrieved from existing data, and an intelligent engineering safety problem investigation and treatment system is realized.

(2) According to the preprocessing method, the deep learning method is introduced into engineering safety management, an intelligent safety problem retrieval mechanism is realized on the premise of ensuring the accuracy of text analysis, the cost of manually processing construction safety management document texts is reduced, and the engineering construction safety management efficiency is improved.

(3) The method and the device calculate the text correlation according to the semantic depth analysis result, and realize intelligent retrieval of the construction safety management text.

Drawings

Fig. 1 schematically shows an exemplary system architecture to which a training method of a document text intelligent retrieval model may be applied according to an embodiment of the present disclosure.

Fig. 2 schematically shows a flowchart of a training method of an engineering construction safety management document text intelligent retrieval model according to an embodiment of the disclosure.

Fig. 3 schematically shows a model block diagram of a training and searching method of an engineering construction safety management document text intelligent searching model according to an embodiment of the disclosure.

FIG. 4 schematically illustrates a twin neural network calculated loss value map according to an embodiment of the present disclosure.

FIG. 5 schematically illustrates a twin neural network computation accuracy graph in accordance with an embodiment of the present disclosure.

Fig. 6 schematically shows a block diagram of a training device of an intelligent text retrieval model for engineering construction safety management documents according to an embodiment of the disclosure.

Fig. 7 schematically shows a Bi-LSTM model structure diagram according to an embodiment of the disclosure.

Fig. 8 schematically shows a block diagram of an intelligent text retrieval device for engineering construction safety management documents according to an embodiment of the present disclosure.

Fig. 9 schematically illustrates a block diagram of a computer system suitable for implementing a training method of an engineering construction safety management document text intelligent retrieval model or an engineering construction safety management document text intelligent retrieval method according to an embodiment of the present disclosure.

Detailed Description

Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood that the description is illustrative only and is not intended to limit the scope of the present disclosure. In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the disclosure. It may be evident, however, that one or more embodiments may be practiced without these specific details. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present disclosure.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.

All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It is noted that the terms used herein should be interpreted as having a meaning that is consistent with the context of this specification and should not be interpreted in an idealized or overly formal sense.

Where a convention analogous to "at least one of A, B and C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B and C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.). Where a convention analogous to "A, B or at least one of C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B or C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.).

With the development of times and technologies, many sudden and high-risk potential safety hazards and safety management problems exist in the construction process of large-scale infrastructure projects, and corresponding solutions need to be found and given in time. The deep semantic analysis improves the efficiency and the accuracy of construction safety management, reduces errors in the construction safety management, enables the construction safety text management to be more intelligent, and provides a theoretical basis for the intelligent management and evaluation of the safety of a construction site.

The embodiment of the disclosure provides a training method and a device for an engineering construction safety management document text intelligent retrieval model, a text retrieval method and a text retrieval device, a computer system and a computer readable storage medium. The training method of the intelligent search model for the engineering construction safety management document text comprises the following steps: collecting sample data; wherein the sample data is characterized by a plurality of text pairs; the text pairs include: a security problem text and a security specification text which are mutually related; dividing the sample data into a training set and a testing set, and preprocessing the training set and the testing set to obtain a safety problem text word vector set and a safety standard text word vector set in the training set and the testing set; respectively inputting the safety problem text word vector set and the safety standard text word vector set in a training set into a first sub-network and a second sub-network of the twin neural network to obtain a safety problem feature vector set and a safety standard text feature vector set, wherein the safety problem feature vector set and the safety standard text feature vector set are used for representing the result of text depth analysis; wherein the first sub-network and the second sub-network of the twin neural network are both generated based on a Bi-LSTM model; determining a correlation value between a safety problem characteristic vector and a safety standard text characteristic vector according to the safety problem characteristic vector set and the safety standard text characteristic vector set, constructing a loss function and calculating a loss value; and respectively inputting the safety problem text word vector set and the safety standard text word vector set in the test set into a first sub-network and a second sub-network of the twin neural network, and performing optimization training on the loss function to obtain the intelligent text retrieval model of the engineering construction safety management document.

Fig. 1 schematically illustrates an exemplary system architecture 100 to which a training method of a document text intelligent retrieval model may be applied, according to an embodiment of the present disclosure. It should be noted that fig. 1 is only an example of a system architecture to which the embodiments of the present disclosure may be applied to help those skilled in the art understand the technical content of the present disclosure, and does not mean that the embodiments of the present disclosure may not be applied to other devices, systems, environments or scenarios.

As shown in fig. 1, the system architecture 100 according to this embodiment may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 serves as a medium for providing communication links between the terminal devices 101, 102, 103 and the server 105. Network 104 may include various connection types, such as wired and/or wireless communication links, and so forth.

The user may use the terminal devices 101, 102, 103 to interact with the server 105 via the network 104 to receive or send messages or the like. The terminal devices 101, 102, 103 may have various communication client applications installed thereon, such as a shopping application, a web browser application, a search application, an instant messaging tool, a mailbox client, and/or social platform software.

The terminal devices 101, 102, 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like.

The server 105 may be a server providing various services, such as a background management server providing text retrieval support for conversations opened by users using the terminal devices 101, 102, 103. The background management server may analyze and otherwise process the received data of the user input sentence, and feed back a processing result (for example, obtain or generate system output data for the user input sentence according to the user input sentence, and the like) to the terminal device.

It should be noted that the training method of the intelligent text retrieval model for engineering construction safety management documents provided by the embodiments of the present disclosure may be generally executed by the server 105. Accordingly, the training device for the intelligent text retrieval model of the engineering construction safety management document provided by the embodiment of the disclosure can be generally arranged in the server 105. The training method for the intelligent text retrieval model of the engineering construction safety management document provided by the embodiment of the disclosure can also be executed by a server or a server cluster which is different from the server 105 and can communicate with the terminal devices 101, 102 and 103 and/or the server 105. Correspondingly, the training device for the intelligent text retrieval model of the engineering construction safety management document provided by the embodiment of the disclosure can also be arranged in a server or a server cluster which is different from the server 105 and can communicate with the terminal devices 101, 102, 103 and/or the server 105. Alternatively, the training method for the intelligent text retrieval model of the engineering construction safety management document provided by the embodiment of the disclosure may also be executed by the terminal device 101, 102, or 103, or may also be executed by another terminal device different from the terminal device 101, 102, or 103. Correspondingly, the training device for the intelligent text retrieval model of the engineering construction safety management document provided by the embodiment of the disclosure can also be arranged in the terminal device 101, 102 or 103, or in other terminal devices different from the terminal device 101, 102 or 103.

For example, the sample data may be originally stored in any of the terminal apparatuses 101, 102, or 103 (e.g., the terminal apparatus 101, but not limited thereto), or stored on an external storage apparatus and may be imported into the terminal apparatus 101. Then, the terminal device 101 may locally execute the training method of the intelligent search model for engineering construction safety management document text provided by the embodiment of the present disclosure, or send the dialog corpus to another terminal device, server, or server cluster, and execute the training method of the intelligent search model for engineering construction safety management document text provided by the embodiment of the present disclosure by another terminal device, server, or server cluster that receives the dialog corpus.

It should be understood that the number of terminal devices, networks, and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.

Fig. 2 schematically shows a flowchart of a training method of an engineering construction safety management document text intelligent retrieval model according to an embodiment of the disclosure.

As shown in fig. 2, the method includes operations S201 to S205.

Operation S201, collecting sample data; wherein the sample data is characterized by a plurality of text pairs; the text pairs include: a security question text and a security specification text associated with each other.

In an embodiment of the present disclosure, the security question text includes: the construction site safety management text, the construction potential safety hazard standing book text and other experience texts collected in the construction site are not listed one by one. The security specification text includes: security management specification standard text or other specification standard text, to name a few.

Operation S202 is performed to divide the sample data into a training set and a test set, and preprocess the training set and the test set to obtain a security problem text word vector set and a security standard text word vector set in the training set and the test set.

In an embodiment of the present disclosure, the operation S202 includes:

dividing the sample data into a training set and a test set according to a set proportion; wherein the training set and the test set include at least one sample data. The specific value of the set ratio is not further limited herein, and may be adjusted as needed, for example, 4: 1.

And performing text Word vector calculation on each safety problem text and each safety standard text in the training set and the test set by using a skip-gram model in the Word2vec technology to obtain a safety problem text Word vector set and a safety standard text Word vector set in the training set and the test set.

In another embodiment of the present disclosure, operation S202 further includes an operation of semantic depth analysis on the basis of the previous embodiment. Specifically, operation S202 includes the following sub-operations:

operation S2021, divide the sample data into a training set and a test set according to a set proportion; wherein the training set and the test set include at least one sample data.

Operation S2022, based on the Jieba word segmentation library, performing Jieba word segmentation on each piece of the safety problem text and the safety standard text in the training set to obtain a plurality of text words in each piece of the safety problem text and the safety standard text.

For example, Jieba word segmentation is performed on the text "potential safety hazard exists", and the text words "existence" and "potential safety hazard" are obtained.

Operation S2023, analyzing the main structure of each safety problem text and each safety standard text in the training set by using a semantic structure model, defining the property of each text word in each safety problem text and each safety standard text, and labeling the text word corresponding to each safety problem text and each safety standard text.

For example, a semantic structure model is used for analyzing the main structure of a text "protective barrier is not provided with a skirting board", a text word "protective barrier" is defined as a subject, a text word "not" is defined as an adverb, a text word "set" is defined as a predicate, and a text word "skirting board" is defined as an object. And are marked on the text words "guard rail", "not", "set" and "skirting board".

Operation S2024, rejecting each text symbol in the safety problem text and the safety specification text in the training set and the test set.

For example, there is a potential safety hazard in the reject text "no guard rail or guard line is set. "text symbol in", "and". ".

Operation S2025, performing text Word vector calculation on each safety problem text and the safety standard text in the training set and the test set by using a skip-gram model in the Word2vec technology to obtain a safety problem text Word vector set and a safety standard text Word vector set in the training set and the test set.

Operation S2026, determining, by using an Attention mechanism, a weight of each text word vector in the safety issue text word vector set and the safety standard text word vector set in the training set and the test set, and blending the word weight into the word vector to obtain the safety issue text word vector and the safety standard text word vector considering the weight for Attention. The specific method of merging may be to multiply the word weight by the word vector.

Operation S203, inputting the safety problem text word vector set and the safety standard text word vector set in a training set into a first subnetwork and a second subnetwork of the twin neural network, respectively, to obtain a safety problem feature vector set and a safety standard text feature vector set, where the safety problem feature vector set and the safety standard text feature vector set are used to represent a result of text depth analysis; wherein the first sub-network and the second sub-network of the twin neural network are both generated based on a Bi-LSTM model. Specifically, operation S203 includes the following sub-operations:

operation S2031, the safety problem text word vector sequentially passes through an input gate, a forgetting gate and an output gate of the Bi-LSTM model corresponding to the first subnetwork in the twin neural network; the safety standard text word vector sequentially passes through an input gate, a forgetting gate and an output gate of the Bi-LSTM model corresponding to the second subnetwork in the twin neural network;

operation S2032, the Bi-LSTM model corresponding to the first sub-network traverses the security problem text word vector along a first direction and a second direction to obtain a security problem feature vector; traversing the safety standard text word vector along a first direction and a second direction by the Bi-LSTM model corresponding to the second sub-network to obtain a safety standard text feature vector; wherein the first direction represents a forward word order direction and the second direction represents a reverse word order direction;

operation S2033 is performed, and the safety problem feature vector set and the safety criterion text feature vector set are obtained by repeating operation S2031 and operation S2032 on each safety problem text word vector in the safety problem text word vector set and each safety criterion text word vector in the safety criterion text word vector set in the training set.

And operation S204, determining a correlation value between the safety problem feature vector and the safety standard text feature vector according to the safety problem feature vector set and the safety standard text feature vector set, constructing a loss function and calculating a loss value. Specifically, operation S204 includes the following sub-operations:

in operation S2041, a manhattan distance is used to calculate a correlation value between the security problem feature vector and the security canonical text feature vector output by the first sub-network and the second sub-network, where the formula is as follows:

D_is(X₁，X₂)＝|x₁-x₂|+|y₁-y₂|

in the formula, X₁、X₂Respectively representing a safety problem feature vector and a safety standard text feature vector; x is the number of₁、y₁Two-dimensional space coordinates of the safety problem feature vector; x is the number of₂、y₂Two-dimensional space coordinates of the text feature vector are specified for safety; dis (X)₁，X₂) Manhattan distance of security problem feature vector and security specification text feature vector, i.e. security problem featureAn association value between the vector and the security specification text feature vector; and

in operation S2042, a contextual Loss function is used to calculate a Loss value, where the formula is as follows:

in the formula: n is the number of the safety problem characteristic vectors and the safety standard text characteristic vectors in the training set; when the sample dissimilarity threshold m is met, the safety problem feature vector is dissimilar to the safety specification text feature vector, and Y is 0; when the sample dissimilarity threshold m is not satisfied, the security problem feature vector is similar to the security specification text feature vector, and Y is 1.

And operation S205, respectively inputting the safety problem text word vector set and the safety specification text word vector set in the test set into the first subnetwork and the second subnetwork of the twin neural network, and performing optimization training on the parameter W in the first subnetwork and the second subnetwork of the twin neural network to obtain the intelligent text retrieval model of the engineering construction safety management document. Wherein, W is a parameter representing the Bi-LSTM model, such as a weight coefficient, a hidden layer dimension and the like.

In an embodiment of the present disclosure, the safety issue text (e.g., construction safety management safety issue text) and the safety specification text (e.g., safety management specification standard text) are processed according to operation S2021 and operation S2022. On the basis of the twin neural network trained in operation S203, the safety problem text in the test set is input into a first sub-network of the twin neural network, and the safety specification text in the test set is input into a second sub-network of the twin neural network structure.

Referring to operation S2041, an associated value of the construction safety management safety problem text and the safety specification text is calculated.

Referring to operation S2042, a loss value is calculated according to the correlation value and the loss function, a correlation determination threshold is set, the security specification text content matched with the security problem is extracted, the correlation size between the texts is output, and intelligent retrieval of the engineering construction security management document text is realized.

According to the above embodiment of the present disclosure, the intelligent search method for the engineering construction safety management document text provided by the present disclosure includes: acquiring an intelligent search model of an engineering construction safety management document text; extracting a safety standard text matched with the input safety problem text through an engineering construction safety management document text intelligent retrieval model, and outputting a correlation value between the safety problem text input by the number of pieces and the safety standard text; the intelligent search model of the engineering construction safety management document text is obtained by training through the training method.

Examples

Fig. 3 schematically shows a model block diagram of a training and searching method of an engineering construction safety management document text intelligent searching model according to an embodiment of the disclosure. Referring to fig. 3, for example, "the distribution box is not hung on the wall", there is a sentence "the distribution box should be hung on the wall" in the test set. Inputting the 'distribution box is not hung on the wall' as a safety problem text word vector to a first sub-network; and inputting the 'distribution box should be hung on the wall' as a safety specification text word vector to a second sub-network, analyzing the text through BI-LSTM, outputting a safety problem characteristic vector corresponding to the 'distribution box should be hung on the wall' by the first sub-network, and outputting a safety specification text characteristic vector corresponding to the 'distribution box should be hung on the wall' by the second sub-network. Because the 'distribution box', 'wall mounting' and 'hanging' exist in the two texts, the structural features are very similar, and the 'not' and 'Yihuan' are used as adverbs and are also similar semantically, the feature vectors output by the Bi-LSTM are also very similar, and finally, the strong relevance of the two sentences is obtained by calculating the Manhattan distance, and a relevance value is output.

In order to verify the accuracy of the intelligent text retrieval method for the engineering construction safety management document based on deep semantic analysis, the whole mathematical modeling process is realized through Python language. The method includes the steps of mainly collecting a certain engineering construction safety management text to obtain 32369 groups of construction safety management data as sample data, recording safety management problems occurring in the construction process and solutions corresponding to the safety management problems, and representing the safety management problem text and the solutions in pairs to obtain a plurality of sample pairs according to operation S201, wherein the sample pairs are shown in table 1.

TABLE 1

According to the sample data in table 1, 25895 training set texts and 6474 test set texts are defined. And quantizing the text data in the table 1 by using a Word2vec technology to obtain a text vector. And inputting the quantized safety management problem text into a first sub-network of the twin neural network, and inputting the quantized safety specification text into a second sub-network of the twin neural network. Setting Bi-LSTM hyper-parameter values, defining the calculation times as 50 times, and finally calculating to obtain the accuracy and the loss rate of the text training, as shown in fig. 4 and 5. According to the calculation result, the calculation accuracy of the twin neural network can reach 83.11%, the accuracy is high, and the calculation requirement is met.

The method comprises the steps of inputting a security problem text into a first sub-network of the twin neural network, sequentially analyzing solutions corresponding to the security problem text in the security specification text, calculating the relevance between the security problem text and the security specification text, setting a relevance threshold value to be 0.75, and outputting the security specification text with the relevance larger than the threshold value and a relevance value, wherein the relevance value is shown in table 2.

TABLE 2

The intelligent retrieval method for the security management document text based on the twin neural network can be obtained through the table 2, can effectively and intelligently search the required text content, and has higher accuracy.

Fig. 6 schematically shows a block diagram of a training device of an intelligent text retrieval model for engineering construction safety management documents according to an embodiment of the disclosure.

As shown in fig. 6, the training device for the intelligent text retrieval model of engineering construction safety management documents comprises: a first acquisition module 610, a pre-processing module 620, a data set construction module 630, a loss function construction module 640, and a training module 650.

A first obtaining module 610, configured to collect sample data; wherein the sample data is characterized by a plurality of text pairs; the text pairs include: a security question text and a security specification text associated with each other.

The preprocessing module 620 is configured to divide the sample data into a training set and a test set, and preprocess the training set and the test set to obtain a security problem text word vector set and a security standard text word vector set in the training set and the test set.

According to an embodiment of the present disclosure, the preprocessing module 620 includes: a configuration unit, a text quantization unit and a focus unit. According to another embodiment of the present disclosure, the difference from the preprocessing module is that the method further includes: a first text processing unit, a second text processing unit and a third text processing unit.

And the configuration unit is used for dividing the sample data into a training set and a test set according to a set proportion. Wherein the training set and the test set include at least one sample data.

The first text processing unit is used for performing Jieba word segmentation on each piece of the safety problem text and the safety standard text in the training set based on a Jieba word segmentation library to obtain a plurality of text words in each piece of the safety problem text and the safety standard text.

And the second text processing unit is used for analyzing the main body structures of each safety problem text and each safety standard text in the training set by utilizing a semantic structure model, defining the properties of each text word in each safety problem text and each safety standard text, and labeling the text words.

And the third text processing unit is used for eliminating text symbols in each safety problem text and each safety standard text in the training set and the test set.

And the text quantization unit is used for performing text Word vector calculation on each safety problem text and the safety standard text in the training set and the test set by using a skip-gram model in the Word2vec technology to obtain a safety problem text Word vector set and a safety standard text Word vector set in the training set and the test set.

And the Attention unit is used for determining the weight of each text word vector in the safety problem text word vector set and the safety standard text word vector set in the training set and the test set by using an Attention mechanism, integrating the word weight into the word vector, and acquiring the safety problem text word vector and the safety standard text word vector considering the weight for Attention.

A data set constructing module 630, configured to input the safety problem text word vector set and the safety standard text word vector set in a training set into a first subnetwork and a second subnetwork of the twin neural network, respectively, to obtain a safety problem feature vector set and a safety standard text feature vector set, where the safety problem feature vector set and the safety standard text feature vector set are used to represent a result of text depth analysis. FIG. 7 is a structural diagram of the Bi-LSTM model. As shown in fig. 7, the first sub-network and the second sub-network of the twin neural network are both generated based on a Bi-LSTM model.

For example, the text entered is "on-site switchbox not wall hung". First, each word in the text is vectorized as an input to a Bi-LSTM-based neural network. And (4) carrying out forward analysis on the text by the Bi-LSTM model corresponding to the neural network. Taking the "switchbox" as an example, the positive LSTM results will be based on location characteristics. It can be found that the left side of the 'distribution box' is 'field' and the right side is 'not', although the effect of the 'distribution box' in the text is shown, the expression information quantity in the 'distribution box' cannot be determined, and the 'distribution box' cannot be judged whether ambiguity exists. At the moment, a text needs to be reversely input, the left side of the distribution box is 'not', the right side of the distribution box is 'field', and therefore specific information of the distribution box in the text is obtained, the distribution box is more accurately positioned, and a feature vector of a whole sentence based on words is obtained and serves as the output of the neural network established based on the Bi-LSTM.

According to an embodiment of the present disclosure, the data set constructing module 630 includes: a first text depth analysis unit and a first traversal unit.

The first text depth analysis unit is used for enabling the safety problem text word vectors to sequentially pass through an input gate, a forgetting gate and an output gate of the Bi-LSTM model corresponding to the first sub-network in the twin neural network; the safety standard text word vector sequentially passes through an input gate, a forgetting gate and an output gate of the Bi-LSTM model corresponding to the second subnetwork in the twin neural network; traversing the safety problem text word vector along a first direction and a second direction by the Bi-LSTM model corresponding to the first sub-network to obtain a safety problem feature vector; traversing the safety standard text word vector along a first direction and a second direction by the Bi-LSTM model corresponding to the second sub-network to obtain a safety standard text feature vector; wherein the first direction represents a forward word order direction and the second direction represents a reverse word order direction; .

And the first traversal unit is used for traversing the safety problem text word vector set and the safety standard text word vector set at the text depth analysis unit to obtain a safety problem feature vector set and a safety standard text feature vector set.

And the loss function constructing module 640 is configured to determine an association value between the security problem feature vector and the security specification text feature vector according to the security problem feature vector set and the security specification text feature vector set, construct a loss function, and calculate a loss value.

According to an embodiment of the present disclosure, the loss function constructing module includes a manhattan distance calculating unit and a loss function unit.

A manhattan distance calculating unit, configured to calculate, by using manhattan distances, association values between the security problem feature vectors and the security specification text feature vectors output by the first sub-network and the second sub-network, where the formula is as follows:

Dis(X₁，X₂)＝|x₁-x₂|+|y₁-y₂|

in the formula, X₁、X₂Respectively representing a safety problem feature vector and a safety standard text feature vector; x is the number of₁、y₁Two-dimensional space coordinates of the safety problem feature vector; x is the number of₂、y₂Two-dimensional space coordinates of the text feature vector are specified for safety; dis (X)₁，X₂) The manhattan distance between the security problem feature vector and the security specification text feature vector is the correlation value between the security problem feature vector and the security specification text feature vector.

The Loss function unit is used for calculating a Loss value by adopting a contrast Loss function, and the formula is as follows:

in the formula: n is the number of the safety problem characteristic vectors and the safety standard text characteristic vectors in the training set; when the sample dissimilarity threshold m is met, the safety problem feature vector is dissimilar to the safety specification text feature vector, and Y is 0; when the sample dissimilarity threshold m is not satisfied, the security problem feature vector is similar to the security specification text feature vector, and Y is 1.

The training module 650 is configured to input the safety problem text word vector set and the safety specification text word vector set in the test set into a first subnetwork and a second subnetwork of the twin neural network respectively, and perform optimization training on the first subnetwork and the second subnetwork of the twin neural network to obtain the intelligent text retrieval model of the engineering construction safety management document.

Fig. 8 schematically shows a block diagram of an intelligent text retrieval device for engineering construction safety management documents according to an embodiment of the present disclosure.

As shown in fig. 8, the intelligent retrieval device for engineering construction safety management document text comprises: a second acquisition module 810 and a retrieval module 820.

And a second obtaining module 810, configured to obtain an intelligent text retrieval model of the engineering construction safety management document. The intelligent search model of the engineering construction safety management document text is obtained by training through the training method of the intelligent search model of the engineering construction safety management document text.

And the retrieval module 820 is used for extracting a safety standard text matched with the input safety problem text through an engineering construction safety management document text intelligent retrieval model and outputting a correlation value between the safety problem text input by the number of pieces and the safety standard text.

Any of the modules, units, sub-units, or at least part of the functionality of any of them according to embodiments of the present disclosure may be implemented in one module. Any one or more of the modules, units and sub-units according to the embodiments of the present disclosure may be implemented by being split into a plurality of modules. Any one or more of the modules, units, sub-units according to the embodiments of the present disclosure may be implemented at least partially as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented in any other reasonable manner of hardware or firmware by integrating or packaging a circuit, or in any one of three implementations of software, hardware, and firmware, or in any suitable combination of any of them. Alternatively, one or more of the modules, units, sub-units according to embodiments of the disclosure may be at least partially implemented as computer program modules, which, when executed, may perform the corresponding functions.

For example, any plurality of the first obtaining module 610, the preprocessing module 620, the data set constructing module 630, the loss function constructing module 640, and the training module 650 may be combined and implemented in one module/unit/sub-unit, or any one of the modules/units/sub-units may be split into a plurality of modules/units/sub-units. Alternatively, at least part of the functionality of one or more of these modules/units/sub-units may be combined with at least part of the functionality of other modules/units/sub-units and implemented in one module/unit/sub-unit. According to an embodiment of the present disclosure, at least one of the first obtaining module 610, the preprocessing module 620, the data set constructing module 630, the loss function constructing module 640, and the training module 650 may be implemented at least partially as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or any other reasonable manner of integrating or packaging a circuit, or any one of three manners of software, hardware, and firmware, or any suitable combination of any of them. Alternatively, at least one of the first obtaining module 610, the preprocessing module 620, the data set construction module 630, the loss function construction module 640 and the training module 650 may be at least partially implemented as a computer program module that, when executed, may perform a corresponding function.

It should be noted that, in the embodiment of the present disclosure, the training device part of the intelligent text retrieval model of the engineering construction safety management document corresponds to the training method part of the intelligent text retrieval model of the engineering construction safety management document in the embodiment of the present disclosure, and the description of the training device part of the intelligent text retrieval model of the engineering construction safety management document specifically refers to the training method part of the intelligent text retrieval model of the engineering construction safety management document, which is not described herein again.

Accordingly, the intelligent text retrieval device part of the engineering construction safety management document in the embodiment of the disclosure corresponds to the intelligent text retrieval method part of the engineering construction safety management document in the embodiment of the disclosure, and the description of the intelligent text retrieval device part of the engineering construction safety management document specifically refers to the intelligent text retrieval method part of the engineering construction safety management document, which is not described herein again.

Fig. 9 schematically illustrates a block diagram of a computer system suitable for implementing a training method of an engineering construction safety management document text intelligent retrieval model or an engineering construction safety management document text intelligent retrieval method according to an embodiment of the present disclosure. The computer system illustrated in FIG. 9 is only one example and should not impose any limitations on the scope of use or functionality of embodiments of the disclosure.

As shown in fig. 9, a computer system 900 according to an embodiment of the present disclosure includes a processor 901 which can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM)902 or a program loaded from a storage section 908 into a Random Access Memory (RAM) 903. Processor 901 may comprise, for example, a general purpose microprocessor (e.g., a CPU), an instruction set processor and/or associated chipset, and/or a special purpose microprocessor (e.g., an Application Specific Integrated Circuit (ASIC)), among others. The processor 901 may also include on-board memory for caching purposes. The processor 901 may comprise a single processing unit or a plurality of processing units for performing the different actions of the method flows according to embodiments of the present disclosure.

In the RAM 903, various programs and data necessary for the operation of the system 900 are stored. The processor 901, the ROM 902, and the RAM 903 are connected to each other through a bus 904. The processor 901 performs various operations of the method flows according to the embodiments of the present disclosure by executing programs in the ROM 902 and/or the RAM 903. Note that the programs may also be stored in one or more memories other than the ROM 902 and the RAM 903. The processor 901 may also perform various operations of the method flows according to embodiments of the present disclosure by executing programs stored in the one or more memories.

System 900 may also include an input/output (I/O) interface 905, input/output (I/O) interface 905 also connected to bus 904, according to an embodiment of the present disclosure. The system 900 may also include one or more of the following components connected to the I/O interface 905: an input portion 906 including a keyboard, a mouse, and the like; an output section 907 including components such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage portion 908 including a hard disk and the like; and a communication section 909 including a network interface card such as a LAN card, a modem, or the like. The communication section 909 performs communication processing via a network such as the internet. The drive 910 is also connected to the I/O interface 905 as necessary. A removable medium 911 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 910 as necessary, so that a computer program read out therefrom is mounted into the storage section 908 as necessary.

According to embodiments of the present disclosure, method flows according to embodiments of the present disclosure may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable storage medium, the computer program containing program code for performing the method illustrated by the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 909, and/or installed from the removable medium 911. The computer program, when executed by the processor 901, performs the above-described functions defined in the system of the embodiment of the present disclosure. The systems, devices, apparatuses, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the present disclosure.

The present disclosure also provides a computer-readable storage medium, which may be contained in the apparatus/device/system described in the above embodiments; or may exist separately and not be assembled into the device/apparatus/system. The computer-readable storage medium carries one or more programs which, when executed, implement the method according to an embodiment of the disclosure.

According to an embodiment of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium. Examples may include, but are not limited to: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

For example, according to embodiments of the present disclosure, a computer-readable storage medium may include the ROM 902 and/or the RAM 903 described above and/or one or more memories other than the ROM 902 and the RAM 903.

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions. Those skilled in the art will appreciate that various combinations and/or combinations of features recited in the various embodiments and/or claims of the present disclosure can be made, even if such combinations or combinations are not expressly recited in the present disclosure. In particular, various combinations and/or combinations of the features recited in the various embodiments and/or claims of the present disclosure may be made without departing from the spirit or teaching of the present disclosure. All such combinations and/or associations are within the scope of the present disclosure.

The embodiments of the present disclosure have been described above. However, these examples are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although the embodiments are described separately above, this does not mean that the measures in the embodiments cannot be used in advantageous combination. The scope of the disclosure is defined by the appended claims and equivalents thereof. Various alternatives and modifications can be devised by those skilled in the art without departing from the scope of the present disclosure, and such alternatives and modifications are intended to be within the scope of the present disclosure.