CN117951547A - Bid and tendered data processing method and device based on artificial intelligence - Google Patents

Abstract

The embodiment of the application provides a bid and ask data processing method and device based on artificial intelligence, wherein the method comprises the following steps: inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bidding document to determine corresponding bidding requirements, and performing model training on a preset decision tree model by taking the bidding requirements as a model training set to obtain a first feature model; determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model; inputting bidding demand features extracted from the first feature model and bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model for feature fusion; the application can effectively improve the matching accuracy and efficiency of bidding documents.

Description

Bid and tendered data processing method and device based on artificial intelligence

Technical Field

The application relates to the field of artificial intelligence, in particular to a bid and ask data processing method and device based on artificial intelligence.

Background

The bidding field is a vital link in commercial transactions, however, a significant problem that currently exists is that the expressions in the bidding documents and in the bidding documents are often not well understood, resulting in inaccuracy and inefficiency in the bidding matching process.

In particular, in bidding documents, the signer is often challenged with difficulty in accurately expressing the requirements. Sometimes, because of improper use of technical terms, ambiguous expressions or insufficient descriptions, bidding documents may cause various explanations and comprehension trouble for bidders. Also, in bid documents, bidders may not adequately meet the needs of the bidders due to inadequate understanding of the bid documents.

In addition, because of the matching of manual auditing bidding documents, the problems of subjective judgment, tedious manual operation and error are existed, resulting in inefficiency of the whole bidding process.

Accordingly, the prior art has a series of problems in writing and matching bidding documents, including but not limited to unclear information, inaccurate matching, inefficiency, etc., and a new technical means is needed to improve the quality and efficiency of bidding process.

The problems are that when the bidding documents are manufactured, the details and potential problems of all aspects need to be considered comprehensively, so that the bidding documents can meet project requirements, the competing principle of fairness notarization can be presented, and efficiency and benefit are considered.

Disclosure of Invention

Aiming at the problems in the prior art, the application provides a bid and ask data processing method and device based on artificial intelligence, which can effectively improve the matching accuracy and efficiency of bid and ask files.

In order to solve at least one of the problems, the application provides the following technical scheme:

In a first aspect, the present application provides an artificial intelligence based bidding data processing method, comprising:

Inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bidding document to determine corresponding bidding requirements, and performing model training on a preset decision tree model by taking the bidding requirements as a model training set to obtain a first feature model;

Determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model;

And inputting the bidding demand features extracted from the first feature model and the bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model for feature fusion, inputting the result of feature fusion into an output layer of the integrated model, and outputting bidding matching data through the output layer of the integrated model.

Further, the step of inputting the preset natural language processing model according to the technical parameters, the service standard, the construction period requirement and the qualification requirement in the historical bidding document to determine the corresponding bidding requirement includes:

Inputting technical parameters, service standards, construction period requirements and qualification requirements in the history bidding documents into a preset deep bidirectional network learning model for pre-training, and determining context-related representation of words;

And adding a bid requirement label to the context-related representation of the word, and outputting a corresponding bid requirement through the deep bidirectional network learning model.

Further, the training the model of the preset decision tree model by using the bid requirement as a model training set to obtain a first feature model includes:

Performing model training on a preset decision tree model by taking the bid-inviting requirement as a model training set and a verification set, gradually increasing the depth value of the decision tree model and/or the minimum sample number required by splitting nodes until the performance of the decision tree model on the verification set starts to be reduced, and determining a corresponding optimal performance depth value;

And obtaining a first characteristic model according to the optimal performance depth value.

Further, the determining the corresponding bidding service according to the team professional description, the technical scheme description and the team qualification input preset natural language processing model in the historical bidding file includes:

Performing word segmentation processing on team professional description, technical scheme description and team qualification in the historical bidding document, adding bidding service labels, and then inputting the obtained results into a preset cyclic neural network model for text classification;

and performing parameter tuning on the cyclic neural network model through a back propagation algorithm to obtain bidding service output by the cyclic neural network model.

Further, the performing model training on the preset cyclic neural network model by using the bidding service as a model training set to obtain a second feature model includes:

determining the maximum length of the text sequence in the bidding service, and filling the text sequence according to the maximum length and a preset zero vector;

Dividing the bidding service filled by the text sequence into a model training set and a verification set, inputting the training set into a preset cyclic neural network model, and performing iterative training on the cyclic neural network model by setting a mean square error loss function to obtain a second characteristic model.

Further, the feature fusion layer for inputting the bidding requirement feature extracted from the first feature model and the bidding service feature extracted from the second feature model into a preset integrated model for feature fusion comprises the following steps:

Extracting a decision tree depth value and node splitting conditions from the first feature model and taking the decision tree depth value and node splitting conditions as bidding demand features, and extracting output of a hidden layer of a circulating neural network and learned time sequence relations from the second feature model as bidding service features;

And inputting the bidding demand features and the bidding service features into a feature fusion layer of a random forest model to perform feature fusion operation.

Further, the inputting the bidding demand features and the bidding service features into a feature fusion layer of a random forest model for feature fusion operation includes:

taking the bidding demand feature as a first feature vector and the bidding service feature as a second feature vector;

And performing splicing operation on the first feature vector and the second feature vector to obtain all feature information from the decision tree model and the cyclic neural network model.

In a second aspect, the present application provides an artificial intelligence based bidding data processing apparatus, comprising:

the first feature model construction module is used for inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bid request file to determine corresponding bid request, and carrying out model training on a preset decision tree model by taking the bid request as a model training set to obtain a first feature model;

The second feature model construction module is used for determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second feature model;

And the model fusion prediction module is used for inputting the bidding demand features extracted from the first feature model and the bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model for feature fusion, inputting the result of the feature fusion into an output layer of the integrated model, and outputting bidding matching data through the output layer of the integrated model.

In a third aspect, the present application provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the artificial intelligence based bidding data processing method when executing the program.

In a fourth aspect, the present application provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the artificial intelligence based bidding data processing method.

In a fifth aspect, the present application provides a computer program product comprising computer programs/instructions which when executed by a processor implement the steps of the artificial intelligence based bidding data processing method.

According to the technical scheme, the application provides an artificial intelligence-based bidding data processing method and device, corresponding bidding requirements are determined by inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bidding document, and the bidding requirements are used as a model training set to carry out model training on a preset decision tree model to obtain a first feature model; determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model; and inputting the bidding demand features extracted from the first feature model and the bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model for feature fusion, so that the matching accuracy and efficiency of bidding documents can be effectively improved.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart of an artificial intelligence based bid and ask data processing method in an embodiment of the application;

FIG. 2 is a second flow chart of an artificial intelligence based bid data processing method in accordance with an embodiment of the present application;

FIG. 3 is a third flow chart of an artificial intelligence based bid and ask data processing method according to an embodiment of the present application;

FIG. 4 is a flow chart of a bid data processing method based on artificial intelligence in an embodiment of the application;

FIG. 5 is a flow chart of an artificial intelligence based bid data processing method in accordance with an embodiment of the present application;

FIG. 6 is a flowchart of an artificial intelligence based bid data processing method according to an embodiment of the present application;

FIG. 7 is a flow chart of a method for processing bid data based on artificial intelligence in accordance with an embodiment of the present application;

FIG. 8 is a block diagram of an artificial intelligence based bidding data processing apparatus in an embodiment of the present application;

fig. 9 is a schematic structural diagram of an electronic device in an embodiment of the application.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.

The technical scheme of the application obtains, stores, uses, processes and the like the data, which all meet the relevant regulations of national laws and regulations.

In consideration of the problems of the prior art in writing and matching bidding documents, including but not limited to problems of unclear information, inaccurate matching, low efficiency and the like, the application provides a bidding data processing method and device based on artificial intelligence, which are used for determining corresponding bidding requirements by inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bidding document, and performing model training on a preset decision tree model by taking the bidding requirements as a model training set to obtain a first characteristic model; determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model; and inputting the bidding required features extracted from the first feature model and bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model to perform feature fusion, so that the matching rate can be effectively improved.

In order to effectively improve the matching accuracy and efficiency of bidding documents, the application provides an embodiment of an artificial intelligence-based bidding data processing method, referring to fig. 1, wherein the artificial intelligence-based bidding data processing method specifically comprises the following steps:

step S101: inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bidding document to determine corresponding bidding requirements, and performing model training on a preset decision tree model by taking the bidding requirements as a model training set to obtain a first feature model;

It will be appreciated that the bidding party may have the following problems when bidding:

1. The core of the bidding documents is to clearly and specifically describe the requirements and targets of the projects, and if the key information such as technical parameters, service standards, construction period requirements and the like of the projects are ambiguous or inaccurate, the bid understanding deviation of bidders can be caused, and the bid quality and the implementation effect of subsequent projects are affected.

2. Unreasonable setting of qualification requirements, and if the qualification conditions of bidders are too severe or do not accord with actual project requirements, enterprises with strength but temporarily without certain specific qualification can be limited to participate in bidding, so that the competitiveness is reduced; conversely, too low a threshold may attract a plurality of enterprises without actual execution capability, and increase screening difficulty.

3. The scoring standard is difficult to formulate, and the scientific, fair and operable scoring standard is a technical activity. How to reasonably distribute the weights of factors such as business, technology, price and the like, which not only needs to ensure fairness but also needs to guide bidders to provide an optimal solution.

Alternatively, the method of Natural Language Processing (NLP) and text semantic recognition may be used in this embodiment to determine the corresponding bid requirement according to the technical parameters, service standards, construction period requirements and qualification requirements in the historical bid document,

Alternatively, the present embodiment may learn the context-dependent word representation through BERT (deep bi-directional transducer network learning model). The history bidding documents can be fine-tuned to enable the history bidding documents to understand semantic relationships among technical parameters, service standards, construction period requirements and qualification requirements.

It is understood that BERT (Bidirectional Encoder Representations from Transformers) is a pre-trained language model that learns context-dependent word representations through a deep bi-directional transducer network. The history bidding documents can be fine-tuned to enable the history bidding documents to understand semantic relationships among technical parameters, service standards, construction period requirements and qualification requirements.

Alternatively, in this embodiment, a data set containing the bid requirement text and corresponding tags (e.g., bid types, service requirements, etc.) may be collected and consolidated. And performing preprocessing operations such as word segmentation, word deactivation and the like on the bidding-required text, and converting the text into an input format acceptable to the model. The tag is then encoded and converted into a format recognizable by the model, such as numerical or unicode. The decision tree model is selected as the model for fitting the bid requirements. The decision tree model can make decisions according to input features and is applicable to classification problems. And taking the bidding demand text as an input characteristic, taking a corresponding label as an output label, and performing model training. After training is completed, a decision tree model for fitting the bid requirements is obtained, which may be referred to as a first feature model.

Through this process, the present embodiment may use the bid requirement text for training the decision tree model, resulting in a first feature model for feature extraction and bid requirement classification. This model can help understand key information in the bidding requirement text, providing valuable features for subsequent tasks.

Step S102: determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model;

it will be appreciated that bidders may have the following problems when writing a bid:

1. The understanding and analysis of the bidding documents are the first part of writing the bidding documents, and include technical specifications, business requirements, legal terms and the like, so that accurate understanding and grasp of the contents are required, and the bidding documents are prevented from being invalidated due to missing key information.

2. Item demand matching degree, which is what is the matching degree, the enterprise needs to accurately evaluate whether its own product or service matches the demand of the bidding item? And how to meet project requirements is clearly set forth in the specification.

3. And the technical scheme design is used for formulating a detailed, feasible and competitive technical solution according to project characteristics and technical requirements, which requires the support of a professional technical team, and ensures the advancement, reliability and economic rationality of the scheme.

4. And preparing qualification proving materials, and collecting and arranging relevant proving materials such as complete enterprise qualification, performance cases, quality certification and the like, so as to ensure that the materials meet the requirements specified by the bidding documents, and any missing or non-compliance can lead to waste bidding.

Alternatively, in this embodiment, in order to determine the corresponding bidding service according to team professional descriptions, technical scheme descriptions and team qualification in the historical bidding document, a Natural Language Processing (NLP) model may be used for implementation.

Optionally, the historical bidding documents are acquired and arranged, including team professional description, technical scheme description, team qualification and other relevant information. Each historical bid file is labeled with a corresponding bid service tag, such as a service type, project size, etc. Text classification is based on models of recurrent neural network RNNs, which can also be used to capture semantic information in text.

Meanwhile, the text can be segmented and labeled, and the text can be converted into an input format acceptable by a model. And filling the text sequence, and ensuring that the lengths of the input texts are consistent so as to adapt to the input requirements of the model. Then, text data of the historical bid files are input into the model, and each file is labeled with a corresponding bid service tag. And performing model fine adjustment on the historical bidding file, and adjusting model parameters through a back propagation algorithm so that semantic relations in the text can be better understood. And inputting a new bidding file by using the trimmed model to predict bidding service.

Alternatively, when training a preset Recurrent Neural Network (RNN) model using the bidding service as a model training set, the present embodiment will obtain a second feature model for feature extraction.

Specifically, bidding services marked in the historical bidding file are used as training sets. Each bidding service text and its corresponding tag (bidding service type) make up a training set sample. The bidding service text is segmented and tokenized, converting the text into an input format acceptable to the model. And filling the text sequence, and ensuring that the lengths of the input texts are consistent so as to adapt to the input requirements of the RNN model. The design of recurrent neural network structures may choose simple RNN structures or more complex variants, such as long short-term memory networks (LSTM) or gated loop units (GRU). An input layer of the RNN model is defined, taking into account the dimensions of the input text. The output layer is defined to ensure that the type of bidding service can be predicted.

Then, data of the bidding service text is input into the RNN model, and each text is labeled with a corresponding bidding service type tag. Through a back propagation algorithm, model parameters are adjusted so that semantic relationships in text can be better understood. And evaluating the trained RNN model by using a verification set, so as to ensure that the model has better generalization performance on unseen data. Depending on the evaluation result, it may be necessary to adjust the hyper-parameters of the model to optimize performance. After training, we have obtained an RNN model for feature extraction, which may be referred to as a second feature model.

And inputting a new bidding file by using the trained RNN model, and extracting the characteristics through the hidden state of the middle layer or the characteristic representation of the output layer.

Step S103: and inputting the bidding demand features extracted from the first feature model and the bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model for feature fusion, inputting the result of feature fusion into an output layer of the integrated model, and outputting bidding matching data through the output layer of the integrated model.

Alternatively, in this embodiment, a trained decision tree model may be used to extract the depth value of the decision tree and the node splitting condition as features of the bid requirement. And extracting the output of the hidden layer and the learned time sequence relationship as the characteristics of the bidding service by using a trained Recurrent Neural Network (RNN) model.

The extracted bid requirement features and bid service features are then input to a feature fusion layer of a random forest model. Random forests are an integrated learning method, which can process multiple features at the same time, and the model performance is improved by combining the prediction results of multiple decision trees. In the feature fusion layer, the random forest model fuses bidding demand features and bidding service features to produce a final feature representation. The output of the random forest model is the fused characteristic representation, which captures the key characteristics of bidding requirements and bidding services. This fused representation of the characteristics may be used for subsequent tasks such as matching bid requirements with bidding services.

Through the process, the feature fusion capability of the random forest model is utilized, the bidding demand features and the bidding service features are integrated together, and the information of the bidding demand features and the bidding service features is captured in a more comprehensive mode, so that the performance of the whole model is improved.

As can be seen from the above description, according to the bid data processing method based on artificial intelligence provided by the embodiment of the present application, the corresponding bid requirement can be determined by inputting the preset natural language processing model according to the technical parameters, the service standard, the construction period requirement and the qualification requirement in the historical bid file, and the bid requirement is used as the model training set to perform model training on the preset decision tree model, so as to obtain the first feature model; determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model; and inputting the bidding required features extracted from the first feature model and bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model to perform feature fusion, so that the matching rate can be effectively improved.

In an embodiment of the artificial intelligence based bidding data processing method of the present application, referring to fig. 2, the following may be further specifically included:

step S201: inputting technical parameters, service standards, construction period requirements and qualification requirements in the history bidding documents into a preset deep bidirectional network learning model for pre-training, and determining context-related representation of words;

step S202: and adding a bid requirement label to the context-related representation of the word, and outputting a corresponding bid requirement through the deep bidirectional network learning model.

Alternatively, the present embodiment may learn the context-dependent word representation through BERT (deep bi-directional transducer network learning model). The history bidding documents can be fine-tuned to enable the history bidding documents to understand semantic relationships among technical parameters, service standards, construction period requirements and qualification requirements.

It is understood that BERT (Bidirectional Encoder Representations from Transformers) is a pre-trained language model that learns context-dependent word representations through a deep bi-directional transducer network. The history bidding documents can be fine-tuned to enable the history bidding documents to understand semantic relationships among technical parameters, service standards, construction period requirements and qualification requirements.

Specifically, BERT employs a two-stage approach of pre-training-fine tuning. In the pre-training phase, the model learns the language model over a large scale of text corpus, thereby obtaining a context-dependent representation of the word. Meanwhile, BERT is a bidirectional language model, and context information on the left side and the right side of a word can be considered simultaneously through a deep bidirectional transducer network, so that the model can better understand the semantics of the word.

Alternatively, after pre-training, the BERT may be tailored to a particular domain or task by fine tuning over the specific task. Such fine-tuning can help the model understand the specific context in the historical bidding documents. The BERT model is loaded into a training environment and data of the historical bid-inviting file is input into the model. By fine tuning on the bidding required tasks, the model parameters are adjusted to adapt to the specific bidding required identification task. And evaluating the performance of the trimmed BERT model by using a verification set, performing model tuning, and ensuring that the model can accurately determine corresponding bid requirements according to technical parameters, service standards, construction period requirements and qualification requirements.

In an embodiment of the artificial intelligence based bidding data processing method of the present application, referring to fig. 3, the following may be further specifically included:

Step S301: performing model training on a preset decision tree model by taking the bid-inviting requirement as a model training set and a verification set, gradually increasing the depth value of the decision tree model and/or the minimum sample number required by splitting nodes until the performance of the decision tree model on the verification set starts to be reduced, and determining a corresponding optimal performance depth value;

step S302: and obtaining a first characteristic model according to the optimal performance depth value.

Optionally, in this embodiment, the bidding requirement text is divided into a training set and a validation set, ensuring that both sets contain enough samples to reflect the overall distribution. The decision tree model is selected as the model for fitting the bid requirements. The decision tree model is initialized, a smaller depth value (the number of layers of the tree) is set, and training of the model is started. A decision tree model is trained on the training set and model performance for the current depth value is evaluated using the validation set.

Optionally, in this embodiment, the depth value of the decision tree is gradually increased, i.e. the number of layers of the tree is increased. The minimum number of samples required to increase node splitting may be considered simultaneously to control tree growth. After each step of increasing depth and node splitting conditions, model performance was evaluated using a validation set. The performance change of the model on the verification set is monitored, and whether the generalization performance of the model is improved along with the increase of the depth value is known. When the performance of the model on the validation set begins to drop, the model is over-fitted and the depth value before that point is selected as the best performance depth value. And applying the performance optimal depth value to the decision tree model to obtain a final first feature model.

In an embodiment of the artificial intelligence based bidding data processing method of the present application, referring to fig. 4, the following may be further specifically included:

Step S401: performing word segmentation processing on team professional description, technical scheme description and team qualification in the historical bidding document, adding bidding service labels, and then inputting the obtained results into a preset cyclic neural network model for text classification;

Step S402: and performing parameter tuning on the cyclic neural network model through a back propagation algorithm to obtain bidding service output by the cyclic neural network model.

Alternatively, in this embodiment, in order to determine the corresponding bidding service according to team professional descriptions, technical scheme descriptions and team qualification in the historical bidding document, a Natural Language Processing (NLP) model may be used for implementation.

Optionally, the historical bidding documents are acquired and arranged, including team professional description, technical scheme description, team qualification and other relevant information. Each historical bid file is labeled with a corresponding bid service tag, such as a service type, project size, etc. Text classification is based on models of recurrent neural network RNNs, which can also be used to capture semantic information in text.

In an embodiment of the artificial intelligence based bidding data processing method of the present application, referring to fig. 5, the following may be further specifically included:

step S501: determining the maximum length of the text sequence in the bidding service, and filling the text sequence according to the maximum length and a preset zero vector;

Step S502: dividing the bidding service filled by the text sequence into a model training set and a verification set, inputting the training set into a preset cyclic neural network model, and performing iterative training on the cyclic neural network model by setting a mean square error loss function to obtain a second characteristic model.

It will be appreciated that the text sequence population is to ensure that the entered text is of consistent length when trained on a Recurrent Neural Network (RNN). In processing text data, due to the difference of different text lengths, the lengths of all text sequences need to be adjusted to be the same through filling so as to meet the input requirement of a model.

Alternatively, the length of the different bid service text may be different, with some shorter and some longer. This variability can lead to the need for ways to handle text of different lengths when constructing an input dataset. This embodiment may select a padding symbol (typically a zero vector) as the padding flag. The maximum length of all text sequences in the dataset is determined. This maximum length may be the length of the longest text of all the texts. For each text sequence, if its length is less than the maximum length, padding is performed with padding symbols at its end until the maximum length is reached.

Alternatively, when training a preset Recurrent Neural Network (RNN) model using the bidding service as a model training set, the present embodiment will obtain a second feature model for feature extraction.

Specifically, bidding services marked in the historical bidding file are used as training sets. Each bidding service text and its corresponding tag (bidding service type) make up a training set sample. The bidding service text is segmented and tokenized, converting the text into an input format acceptable to the model. And filling the text sequence, and ensuring that the lengths of the input texts are consistent so as to adapt to the input requirements of the RNN model. The design of recurrent neural network structures may choose simple RNN structures or more complex variants, such as long short-term memory networks (LSTM) or gated loop units (GRU). An input layer of the RNN model is defined, taking into account the dimensions of the input text. The output layer is defined to ensure that the type of bidding service can be predicted.

Then, data of the bidding service text is input into the RNN model, and each text is labeled with a corresponding bidding service type tag. Through a back propagation algorithm, model parameters are adjusted so that semantic relationships in text can be better understood. And evaluating the trained RNN model by using a verification set, so as to ensure that the model has better generalization performance on unseen data. Depending on the evaluation result, it may be necessary to adjust the hyper-parameters of the model to optimize performance. After training, we have obtained an RNN model for feature extraction, which may be referred to as a second feature model.

In an embodiment of the artificial intelligence based bidding data processing method of the present application, referring to fig. 6, the following may be further specifically included:

Step S601: extracting a decision tree depth value and node splitting conditions from the first feature model and taking the decision tree depth value and node splitting conditions as bidding demand features, and extracting output of a hidden layer of a circulating neural network and learned time sequence relations from the second feature model as bidding service features;

step S602: and inputting the bidding demand features and the bidding service features into a feature fusion layer of a random forest model to perform feature fusion operation.

Alternatively, in this embodiment, a trained decision tree model may be used to extract the depth value of the decision tree and the node splitting condition as features of the bid requirement. And extracting the output of the hidden layer and the learned time sequence relationship as the characteristics of the bidding service by using a trained Recurrent Neural Network (RNN) model.

The extracted bid requirement features and bid service features are then input to a feature fusion layer of a random forest model. Random forests are an integrated learning method, which can process multiple features at the same time, and the model performance is improved by combining the prediction results of multiple decision trees. In the feature fusion layer, the random forest model fuses bidding demand features and bidding service features to produce a final feature representation. The output of the random forest model is the fused characteristic representation, which captures the key characteristics of bidding requirements and bidding services. This fused representation of the characteristics may be used for subsequent tasks such as matching bid requirements with bidding services.

In an embodiment of the artificial intelligence based bidding data processing method of the present application, referring to fig. 7, the following may be further specifically included:

step S701: taking the bidding demand feature as a first feature vector and the bidding service feature as a second feature vector;

Step S702: and performing splicing operation on the first feature vector and the second feature vector to obtain all feature information from the decision tree model and the cyclic neural network model.

Alternatively, in this embodiment, the decision tree depth value and the node splitting condition may be extracted from the first feature model to form the bid-requiring feature vector. And extracting the output of the hidden layer of the recurrent neural network and the learned time sequence relationship from the second feature model to form the bidding service feature vector. Then, the first feature vector and the second feature vector are spliced together to form a composite feature vector. The spliced comprehensive feature vector contains all important feature information from the decision tree model and the cyclic neural network model.

In order to effectively improve the matching accuracy and efficiency of bidding documents, the present application provides an embodiment of an artificial intelligence based bidding data processing apparatus for implementing all or part of the content of the artificial intelligence based bidding data processing method, referring to fig. 8, the artificial intelligence based bidding data processing apparatus specifically includes the following contents:

The first feature model construction module 10 is configured to input a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bid amount file to determine corresponding bid amount requirements, and perform model training on a preset decision tree model by using the bid amount requirements as a model training set to obtain a first feature model;

The second feature model construction module 20 is configured to determine a corresponding bidding service according to the team professional description, the technical scheme description and the team qualification input preset natural language processing model in the historical bidding file, and perform model training on the preset cyclic neural network model by using the bidding service as a model training set to obtain a second feature model;

The model fusion prediction module 30 is configured to input the bidding demand feature extracted from the first feature model and the bidding service feature extracted from the second feature model into a feature fusion layer of a preset integrated model for feature fusion, input a result of the feature fusion into an output layer of the integrated model, and output bidding matching data through the output layer of the integrated model.

As can be seen from the above description, the bid-bidding data processing apparatus based on artificial intelligence provided by the embodiment of the present application can determine the corresponding bid-bidding requirement by inputting the preset natural language processing model according to the technical parameters, the service standard, the construction period requirement and the qualification requirement in the historical bid-bidding document, and perform model training on the preset decision tree model by using the bid-bidding requirement as the model training set to obtain the first feature model; determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model; and inputting the bidding required features extracted from the first feature model and bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model to perform feature fusion, so that the matching rate can be effectively improved.

In order to effectively improve matching accuracy and efficiency of bidding documents from a hardware level, the application provides an embodiment of an electronic device for implementing all or part of contents in the artificial intelligence-based bidding data processing method, wherein the electronic device specifically comprises the following contents:

A processor (processor), a memory (memory), a communication interface (Communications Interface), and a bus; the processor, the memory and the communication interface complete communication with each other through the bus; the communication interface is used for realizing information transmission between the bid and ask data processing device based on artificial intelligence and related equipment such as a core service system, a user terminal, a related database and the like; the logic controller may be a desktop computer, a tablet computer, a mobile terminal, etc., and the embodiment is not limited thereto. In this embodiment, the logic controller may refer to the embodiment of the bid data processing method based on artificial intelligence in the embodiment and the embodiment of the bid data processing device based on artificial intelligence, and the contents thereof are incorporated herein and are not repeated here.

It is understood that the user terminal may include a smart phone, a tablet electronic device, a network set top box, a portable computer, a desktop computer, a Personal Digital Assistant (PDA), a vehicle-mounted device, a smart wearable device, etc. Wherein, intelligent wearing equipment can include intelligent glasses, intelligent wrist-watch, intelligent bracelet etc..

In practical applications, part of the bid and ask data processing method based on artificial intelligence can be executed on the side of the electronic device as described above, or all operations can be completed in the client device. Specifically, the selection may be made according to the processing capability of the client device, and restrictions of the use scenario of the user. The application is not limited in this regard. If all operations are performed in the client device, the client device may further include a processor.

The client device may have a communication module (i.e. a communication unit) and may be connected to a remote server in a communication manner, so as to implement data transmission with the server. The server may include a server on the side of the task scheduling center, and in other implementations may include a server of an intermediate platform, such as a server of a third party server platform having a communication link with the task scheduling center server. The server may include a single computer device, a server cluster formed by a plurality of servers, or a server structure of a distributed device.

Fig. 9 is a schematic block diagram of a system configuration of an electronic device 9600 according to an embodiment of the present application. As shown in fig. 9, the electronic device 9600 may include a central processor 9100 and a memory 9140; the memory 9140 is coupled to the central processor 9100. Notably, this fig. 9 is exemplary; other types of structures may also be used in addition to or in place of the structures to implement telecommunications functions or other functions.

In one embodiment, the artificial intelligence based bidding data processing method functionality may be integrated into the central processor 9100. The central processor 9100 may be configured to perform the following control:

step S101: inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bidding document to determine corresponding bidding requirements, and performing model training on a preset decision tree model by taking the bidding requirements as a model training set to obtain a first feature model;

Step S102: determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model;

step S103: and inputting the bidding demand features extracted from the first feature model and the bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model for feature fusion, inputting the result of feature fusion into an output layer of the integrated model, and outputting bidding matching data through the output layer of the integrated model.

As can be seen from the above description, in the electronic device provided by the embodiment of the present application, the corresponding bid requirement is determined by inputting the preset natural language processing model according to the technical parameters, the service standard, the construction period requirement and the qualification requirement in the historical bid document, and the bid requirement is used as a model training set to perform model training on the preset decision tree model, so as to obtain the first feature model; determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model; and inputting the bidding required features extracted from the first feature model and bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model to perform feature fusion, so that the matching rate can be effectively improved.

In another embodiment, the artificial intelligence based bidding data processing apparatus may be configured separately from the central processor 9100, for example, the artificial intelligence based bidding data processing apparatus may be configured as a chip connected to the central processor 9100, and the artificial intelligence based bidding data processing method functions are implemented by control of the central processor.

As shown in fig. 9, the electronic device 9600 may further include: a communication module 9110, an input unit 9120, an audio processor 9130, a display 9160, and a power supply 9170. It is noted that the electronic device 9600 need not include all of the components shown in fig. 9; in addition, the electronic device 9600 may further include components not shown in fig. 9, and reference may be made to the related art.

As shown in fig. 9, the central processor 9100, sometimes referred to as a controller or operational control, may include a microprocessor or other processor device and/or logic device, which central processor 9100 receives inputs and controls the operation of the various components of the electronic device 9600.

The memory 9140 may be, for example, one or more of a buffer, a flash memory, a hard drive, a removable media, a volatile memory, a non-volatile memory, or other suitable device. The information about failure may be stored, and a program for executing the information may be stored. And the central processor 9100 can execute the program stored in the memory 9140 to realize information storage or processing, and the like.

The input unit 9120 provides input to the central processor 9100. The input unit 9120 is, for example, a key or a touch input device. The power supply 9170 is used to provide power to the electronic device 9600. The display 9160 is used for displaying display objects such as images and characters. The display may be, for example, but not limited to, an LCD display.

The memory 9140 may be a solid state memory such as Read Only Memory (ROM), random Access Memory (RAM), SIM card, etc. But also a memory which holds information even when powered down, can be selectively erased and provided with further data, an example of which is sometimes referred to as EPROM or the like. The memory 9140 may also be some other type of device. The memory 9140 includes a buffer memory 9141 (sometimes referred to as a buffer). The memory 9140 may include an application/function storage portion 9142, the application/function storage portion 9142 storing application programs and function programs or a flow for executing operations of the electronic device 9600 by the central processor 9100.

The memory 9140 may also include a data store 9143, the data store 9143 for storing data, such as contacts, digital data, pictures, sounds, and/or any other data used by an electronic device. The driver storage portion 9144 of the memory 9140 may include various drivers of the electronic device for communication functions and/or for performing other functions of the electronic device (e.g., messaging applications, address book applications, etc.).

The communication module 9110 is a transmitter/receiver 9110 that transmits and receives signals via an antenna 9111. A communication module (transmitter/receiver) 9110 is coupled to the central processor 9100 to provide input signals and receive output signals, as in the case of conventional mobile communication terminals.

Based on different communication technologies, a plurality of communication modules 9110, such as a cellular network module, a bluetooth module, and/or a wireless local area network module, etc., may be provided in the same electronic device. The communication module (transmitter/receiver) 9110 is also coupled to a speaker 9131 and a microphone 9132 via an audio processor 9130 to provide audio output via the speaker 9131 and to receive audio input from the microphone 9132 to implement usual telecommunications functions. The audio processor 9130 can include any suitable buffers, decoders, amplifiers and so forth. In addition, the audio processor 9130 is also coupled to the central processor 9100 so that sound can be recorded locally through the microphone 9132 and sound stored locally can be played through the speaker 9131.

An embodiment of the present application further provides a computer-readable storage medium capable of implementing all steps in the artificial intelligence-based bidding data processing method in which the execution subject in the above embodiment is a server or a client, the computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements all steps in the artificial intelligence-based bidding data processing method in which the execution subject in the above embodiment is a server or a client, for example, the processor implements the following steps when executing the computer program:

step S101: inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bidding document to determine corresponding bidding requirements, and performing model training on a preset decision tree model by taking the bidding requirements as a model training set to obtain a first feature model;

Step S102: determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model;

step S103: and inputting the bidding demand features extracted from the first feature model and the bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model for feature fusion, inputting the result of feature fusion into an output layer of the integrated model, and outputting bidding matching data through the output layer of the integrated model.

As can be seen from the above description, the computer readable storage medium provided by the embodiment of the present application determines the corresponding bid requirement by inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bid file, and performs model training on a preset decision tree model by using the bid requirement as a model training set to obtain a first feature model; determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model; and inputting the bidding required features extracted from the first feature model and bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model to perform feature fusion, so that the matching rate can be effectively improved.

Embodiments of the present application also provide a computer program product capable of implementing all the steps in the artificial intelligence based bidding data processing method in which the execution subject in the above embodiments is a server or a client, the computer program/instructions implementing the steps of the artificial intelligence based bidding data processing method when executed by a processor, for example, the computer program/instructions implementing the steps of:

step S101: inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bidding document to determine corresponding bidding requirements, and performing model training on a preset decision tree model by taking the bidding requirements as a model training set to obtain a first feature model;

Step S102: determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model;

step S103: and inputting the bidding demand features extracted from the first feature model and the bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model for feature fusion, inputting the result of feature fusion into an output layer of the integrated model, and outputting bidding matching data through the output layer of the integrated model.

As can be seen from the above description, the computer program product provided by the embodiment of the present application determines the corresponding bid requirement by inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bid document, and performs model training on a preset decision tree model by using the bid requirement as a model training set to obtain a first feature model; determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model; and inputting the bidding required features extracted from the first feature model and bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model to perform feature fusion, so that the matching rate can be effectively improved.

It will be apparent to those skilled in the art that embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (devices), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The principles and embodiments of the present invention have been described in detail with reference to specific examples, which are provided to facilitate understanding of the method and core ideas of the present invention; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present invention, the present description should not be construed as limiting the present invention in view of the above.

Claims (8)

1. An artificial intelligence-based bid and ask data processing method, characterized in that the method comprises the following steps:

Inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bidding document to determine corresponding bidding requirements, and performing model training on a preset decision tree model by taking the bidding requirements as a model training set to obtain a first feature model;

Determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second characteristic model;

And inputting the bidding demand features extracted from the first feature model and the bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model for feature fusion, inputting the result of feature fusion into an output layer of the integrated model, and outputting bidding matching data through the output layer of the integrated model.

2. The method for processing bid data based on artificial intelligence according to claim 1, wherein the determining the corresponding bid requirement according to the technical parameters, the service standard, the construction period requirement and the qualification requirement in the history bid document input the preset natural language processing model comprises:

Inputting technical parameters, service standards, construction period requirements and qualification requirements in the history bidding documents into a preset deep bidirectional network learning model for pre-training, and determining context-related representation of words;

And adding a bid requirement label to the context-related representation of the word, and outputting a corresponding bid requirement through the deep bidirectional network learning model.

3. The artificial intelligence based bidding data processing method according to claim 1, wherein the performing model training on a preset decision tree model with the bidding requirement as a model training set to obtain a first feature model includes:

Performing model training on a preset decision tree model by taking the bid-inviting requirement as a model training set and a verification set, gradually increasing the depth value of the decision tree model and/or the minimum sample number required by splitting nodes until the performance of the decision tree model on the verification set starts to be reduced, and determining a corresponding optimal performance depth value;

And obtaining a first characteristic model according to the optimal performance depth value.

4. The artificial intelligence based bidding data processing method according to claim 1, wherein the determining the corresponding bidding service according to the team professional description, the technical scheme description and the team qualification input preset natural language processing model in the historical bidding file comprises:

Performing word segmentation processing on team professional description, technical scheme description and team qualification in the historical bidding document, adding bidding service labels, and then inputting the obtained results into a preset cyclic neural network model for text classification;

and performing parameter tuning on the cyclic neural network model through a back propagation algorithm to obtain bidding service output by the cyclic neural network model.

5. The artificial intelligence based bidding data processing method according to claim 1, wherein the performing model training on a preset cyclic neural network model with the bidding service as a model training set to obtain a second feature model comprises:

determining the maximum length of the text sequence in the bidding service, and filling the text sequence according to the maximum length and a preset zero vector;

Dividing the bidding service filled by the text sequence into a model training set and a verification set, inputting the training set into a preset cyclic neural network model, and performing iterative training on the cyclic neural network model by setting a mean square error loss function to obtain a second characteristic model.

6. The artificial intelligence based bidding data processing method of claim 1, wherein the feature fusion layer that inputs bidding demand features extracted from the first feature model and bidding service features extracted from the second feature model into a preset integration model performs feature fusion, comprising:

Extracting a decision tree depth value and node splitting conditions from the first feature model and taking the decision tree depth value and node splitting conditions as bidding demand features, and extracting output of a hidden layer of a circulating neural network and learned time sequence relations from the second feature model as bidding service features;

And inputting the bidding demand features and the bidding service features into a feature fusion layer of a random forest model to perform feature fusion operation.

7. The artificial intelligence based bidding data processing method of claim 6, wherein the inputting the bidding demand feature and the bidding service feature into a feature fusion layer of a random forest model for feature fusion operation comprises:

taking the bidding demand feature as a first feature vector and the bidding service feature as a second feature vector;

And performing splicing operation on the first feature vector and the second feature vector to obtain all feature information from the decision tree model and the cyclic neural network model.

8. An artificial intelligence based bidding data processing apparatus, the apparatus comprising:

the first feature model construction module is used for inputting a preset natural language processing model according to technical parameters, service standards, construction period requirements and qualification requirements in a historical bid request file to determine corresponding bid request, and carrying out model training on a preset decision tree model by taking the bid request as a model training set to obtain a first feature model;

The second feature model construction module is used for determining corresponding bidding services according to team professional descriptions, technical scheme descriptions and team qualification input preset natural language processing models in the historical bidding files, and performing model training on a preset cyclic neural network model by taking the bidding services as a model training set to obtain a second feature model;

And the model fusion prediction module is used for inputting the bidding demand features extracted from the first feature model and the bidding service features extracted from the second feature model into a feature fusion layer of a preset integrated model for feature fusion, inputting the result of the feature fusion into an output layer of the integrated model, and outputting bidding matching data through the output layer of the integrated model.