Intelligent demand priority assignment system based on demand semantic analysis

Abstract

The invention provides an intelligent demand priority assignment system based on demand semantic analysis, relates to the field of intelligent demand management, and is characterized in that semantic analysis is performed on a first demand text and a second demand text by using a natural language analysis and coding technology based on artificial intelligence, meanwhile, a business rule is extracted and subjected to semantic analysis, and then, the first demand text and the second demand text are subjected to semantic analysis; calculating a demand difference semantic analysis feature between the first demand text semantic analysis feature and the second demand text semantic analysis feature, therefore, the priority of the first demand and the priority of the second demand are intelligently assigned according to bidirectional attention demand difference semantic representation based on business rule guidance between the demand difference semantic analysis features and the business rule semantic analysis features. In this way, the accuracy and rationality of demand priority evaluation can be effectively improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent demand management, and more specifically, to a demand priority intelligent assignment system based on demand semantic analysis. Background Art

[0002] With the rapid development of digitalization, enterprises and organizations are facing massive demands, covering software development, business process optimization, product iteration and other aspects. Efficient demand management is the key to ensuring project success and sustainable business development. By determining demand priorities, the allocation and assignment of limited resources can be optimized to ensure that the most important demands are handled in a timely manner, thereby improving project success rate and resource utilization efficiency.

[0003] However, traditional methods of determining demand priorities, such as relying on experience, simple voting, or ranking by customer importance, are no longer able to meet the complex and ever-changing business needs of today. These methods often lack scientificity and objectivity, and are easily interfered by subjective factors, resulting in inaccurate demand priority judgments, which in turn affects the progress, cost, and quality of the project. In addition, modern business rules are often complex and changeable, and traditional technologies face many challenges in extracting and applying these rules. It is difficult to effectively combine business rules with requirements, and thus it is impossible to ensure that priority assignments meet actual business needs.

[0004] Therefore, an intelligent demand priority assignment system based on demand semantic analysis is desired. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a demand priority intelligent assignment system based on demand semantic analysis.

[0006] According to one aspect of the present application, a system for intelligently assigning demand priorities based on demand semantic parsing is provided, which includes: a demand text acquisition module, used to acquire a first demand text and a second demand text; a demand text semantic parsing module, used to perform semantic parsing on the first demand text and the second demand text to obtain a first demand text semantic parsing encoding vector and a second demand text semantic parsing encoding vector; a business rule semantic parsing module, used to extract business rules and perform semantic parsing on the business rules to obtain a business rule semantic parsing encoding vector; a demand difference calculation module, used to calculate a demand difference semantic parsing encoding vector between the first demand text semantic parsing encoding vector and the second demand text semantic parsing encoding vector; a demand difference bidirectional evaluation module, used to perform a demand difference bidirectional attention evaluation based on business rules on the demand difference semantic parsing encoding vector and the business rule semantic parsing encoding vector to obtain a demand difference evaluation semantic encoding vector under business rules; and an assignment result generation module, used to obtain an assignment result based on the demand difference evaluation semantic encoding vector under the business rules.

[0007] Compared with the prior art, the demand priority intelligent assignment system based on demand semantic analysis provided by the present application uses natural language analysis and coding technology based on artificial intelligence to perform semantic analysis on the first demand text and the second demand text, and at the same time, extracts business rules and performs semantic analysis on them, and then calculates the demand difference semantic analysis features between the semantic analysis features of the first demand text and the semantic analysis features of the second demand text, so as to intelligently assign the priority of the first demand and the second demand based on the two-way attention demand difference semantic representation guided by business rules between the demand difference semantic analysis features and the business rule semantic analysis features. In this way, the accuracy and rationality of demand priority evaluation can be effectively improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0009] Figure 1 4 is a system block diagram of a system for intelligently assigning demand priorities based on demand semantic analysis according to an embodiment of the present application.

[0010] Figure 2 Schematic diagram of data flow of a demand priority intelligent assignment system based on demand semantic analysis according to an embodiment of the present application.

[0011] Figure 3It is a block diagram of a demand difference bidirectional evaluation module in a demand priority intelligent assignment system based on demand semantic analysis according to an embodiment of the present application. DETAILED DESCRIPTION

[0012] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0013] In today's era of rapid digital development, enterprises and organizations are facing multiple demand challenges, including software development, business process optimization, and product iteration. Efficient demand management has become a core element to ensure project success and sustainable business development. By scientifically determining the priority of demand, limited resources can be more reasonably allocated and utilized, ensuring that key demands can be responded to in a timely manner, thereby improving the success rate of projects and resource utilization efficiency.

[0014] However, traditional methods of determining demand priorities, such as relying on experience, simple voting, or ranking based solely on customer importance, are no longer able to adapt to today's complex and ever-changing business environment. These methods often lack objectivity and scientific basis, and are easily affected by subjective factors, which can lead to deviations in priority judgments, which in turn have an adverse impact on project schedules, costs, and quality. In addition, modern business rules are becoming increasingly complex and changing rapidly, and traditional technologies face many difficulties in extracting and applying these rules. It is difficult to effectively combine the determination of demand priorities with actual business rules, and ultimately it is impossible to ensure that the assignment of priorities can truly meet business needs.

[0015] In response to the above technical problems, the technical concept of the present application is to obtain the first requirement text and the second requirement text, use artificial intelligence-based natural language analysis and coding technology to perform semantic analysis on the first requirement text and the second requirement text, and at the same time, extract business rules and perform semantic analysis on them, and then calculate the semantic analysis features of the demand difference between the semantic analysis features of the first requirement text and the semantic analysis features of the second requirement text, so as to intelligently assign the priority of the first requirement and the second requirement based on the two-way attention demand difference semantic representation guided by business rules between the semantic analysis features of the demand difference and the semantic analysis features of the business rules. The present application can perform in-depth semantic analysis of the demand text to ensure a more accurate and comprehensive understanding of the demand content and reduce the influence of subjective judgment. In addition, the business rules are closely combined with the demand difference assessment, and the importance weight of the business rules is fully considered, thereby improving the accuracy and rationality of the assessment.

[0016] Figure 1 4 is a system block diagram of a system for intelligently assigning demand priorities based on demand semantic analysis according to an embodiment of the present application. Figure 2FIG. 1 is a data flow diagram of a demand priority intelligent assignment system based on demand semantic analysis according to an embodiment of the present application. Figure 1 and Figure 2 As shown, in the demand priority intelligent assignment system 100 based on demand semantic analysis, it includes: a demand text acquisition module 110, which is used to acquire a first demand text and a second demand text; a demand text semantic analysis module 120, which is used to perform semantic analysis on the first demand text and the second demand text to obtain a first demand text semantic analysis coding vector and a second demand text semantic analysis coding vector; a business rule semantic analysis module 130, which is used to extract business rules and perform semantic analysis on the business rules to obtain business rule semantic analysis coding vectors; a demand difference calculation module 140, which is used to calculate the demand difference semantic analysis coding vector between the first demand text semantic analysis coding vector and the second demand text semantic analysis coding vector; a demand difference bidirectional evaluation module 150, which is used to perform a demand difference bidirectional attention evaluation based on business rules on the demand difference semantic analysis coding vector and the business rule semantic analysis coding vector to obtain a demand difference evaluation semantic coding vector under business rules; and an assignment result generation module 160, which is used to obtain an assignment result based on the demand difference evaluation semantic coding vector under the business rules.

[0017] In an embodiment of the present application, the requirement text acquisition module 110 is used to obtain the first requirement text and the second requirement text. It should be understood that the first requirement text and the second requirement text will contain detailed descriptions of specific business requirements, value information that can be brought to the business after completing this requirement, time sensitivity, dependencies with other business requirements, etc. By performing semantic analysis on the content in the first requirement text and the second requirement text, the model can understand the business value, scope of impact, urgency, etc. of different requirements, determine the priority of requirements based on this information, and make more scientific and objective priority decisions. In particular, the first requirement text and the second requirement text can be obtained from project requirement documents and customer requirement records.

[0018] In an embodiment of the present application, the requirement text semantic parsing module 120 is used to perform semantic parsing on the first requirement text and the second requirement text to obtain a first requirement text semantic parsing encoding vector and a second requirement text semantic parsing encoding vector. Specifically, in an embodiment of the present application, the requirement text semantic parsing module is used to: use a semantic encoder including a Bert model to perform semantic parsing on the first requirement text and the second requirement text to obtain a first requirement text semantic parsing encoding vector and a second requirement text semantic parsing encoding vector. It should be understood that the first requirement text and the second requirement text contain rich and diverse semantic information, and this information has a close semantic association between the context. Based on this, in the technical solution of the present application, the first requirement text and the second requirement text are semantically parsed using a semantic encoder including a Bert model to capture the complex semantic information and contextual dependencies in the text, and obtain a first requirement text semantic parsing encoding vector and a second requirement text semantic parsing encoding vector. A person of ordinary skill in the art should know that the Bert model is a pre-trained language model, which is based on the Transformer architecture and introduces a bidirectional encoding mechanism. Unlike traditional unidirectional language models, the Bert model can consider the left and right context information of a word at the same time, so as to more accurately understand the true meaning of each word in the sentence. In other words, the semantics of demand texts are complex and contextual. The Bert model uses a multi-head attention mechanism to consider the information of other words when encoding each word, so as to accurately understand the semantics of the sentence. In this way, the obtained vector can fully reflect the true meaning of the demand text and its internal logic, laying the foundation for further intelligent analysis.

[0019] The following is a detailed description of a specific implementation process of "using a semantic encoder including a Bert model to semantically parse the first requirement text and the second requirement text to obtain the semantic parsing encoding vector of the first requirement text and the semantic parsing encoding vector of the second requirement text": First, it is necessary to choose from a number of pre-trained versions of the Bert model, such as Bert-base-uncased and Bert-large-uncased. These versions are carefully trained on massive amounts of text data, have accumulated rich semantic information, and have strong language understanding capabilities. However, the specific version to be selected needs to be determined based on actual needs and computing resource conditions. For example, if a relatively conventional requirement text is being processed and computing resources are limited, Bert-base-uncased may be a good choice; for more complex situations that require higher semantic understanding accuracy, Bert-large-uncased may be more competent.

[0020] Next is the text preprocessing stage. The first step in this stage is word segmentation, which is an important operation to convert the text into a form that can be processed by the model. Generally, a word segmenter that matches the selected Bert model is used to process the first demand text and the second demand text separately. For the first demand text, assuming it is a more complex description, such as "a new online shopping platform needs to be developed. The platform should have a convenient search function. Users can quickly find the products they want through a variety of screening conditions. It also supports multiple payment methods and has an intuitive user interface for easy user operation." The word segmenter will split it into basic elements. It may split this sentence into words or sub-word units such as ["need", "develop", "a", "new", "online", "shopping", "platform", "," "the", "platform", "should", "have", "convenient", "search", "function", "," users", "can", "through", "multiple", "screening", "conditions", "quickly", "find", "the", "wanted", "products", "," "at the same time", "support", "multiple", "payment", "methods", "," and", "must have", "an", "intuitive", "user", "interface", "," convenient", "user", "operation"], and even more fine-grained split into sub-words, because the input of the Bert model is based on its pre-trained vocabulary, and this split can ensure that the model can accurately process each element in the text. For the second demand text, no matter what its specific content is, it will also undergo the same word segmentation operation.

[0021] After word segmentation, special tags need to be added to these texts. This is a requirement of the Bert model input format. A [CLS] tag will be added at the beginning of each demand text. This tag indicates the beginning of the sentence, allowing the Bert model to identify the starting position of the text, and a [SEP] tag will be added at the end to indicate the end of the sentence or to separate different sentences. For the first demand text, it will become [CLS] A new online shopping platform needs to be developed. The platform should have a convenient search function. Users can quickly find the products they want through a variety of filtering conditions, support multiple payment methods, and have an intuitive user interface that is convenient for users to operate [SEP]. The same operation will be performed for the second demand text. In addition, if it involves a demand text with multiple sentences, different sentences will be separated by [SEP] tags, so that the Bert model can clearly distinguish different sentence structures.

[0022] In addition to special tags, position encoding and fragment encoding are also indispensable parts of preprocessing. Position encoding plays a key role in the model's understanding of the position information of words in a sentence. It adds a code to each word indicating its position, so that the model can know the order of words. For example, in the first requirement text mentioned above, "need" may be assigned a position code, "development" may be assigned another position code, and so on. Words in different positions will have different position codes, allowing the model to distinguish the order of words, because in natural language, the order of words is crucial to semantic understanding. Fragment encoding is to distinguish different sentences. When the first requirement text and the second requirement text are used as input, they will be marked as belonging to different fragments, which helps the Bert model distinguish different input parts.

[0023] After completing these preprocessing tasks, the processed first requirement text and second requirement text can be input into the Bert model. The processing process of the Bert model is an intelligent multi-layer architecture processing flow. First, in the word embedding layer, it maps the input word tags to vector representations. This means that each word tag will find the corresponding initial vector representation based on Bert's pre-trained vocabulary, and this vector contains the original semantic information of the word. Then, these initial vectors will be added to the position encoding and fragment encoding vectors to form a comprehensive input vector. Such a comprehensive vector not only contains the semantics of the word, but also incorporates the position information of the word in the sentence and the fragment information of the sentence, providing a richer information basis for subsequent processing.

[0024] The input vector then enters the multi-layer Transformer architecture. This is the core part of the Bert model, and the multi-head self-attention mechanism is its key feature. In this process, each word will pay attention to the information of other words in the sentence through the multi-head self-attention mechanism, and will also be paid attention to by other words. Taking the first demand text as an example, for the word "search function", during the processing, it will establish a connection with other words such as "convenience", "filter conditions", and "users" because there is a semantic connection between them. The vector representation of "search function" will be dynamically adjusted according to its context in the sentence, rather than viewing the word in isolation. For example, in the part "users can quickly find the products they want through a variety of filter conditions", the semantics of "search function" will be affected by words such as "filter conditions" and "find products". Through the multi-head self-attention mechanism, "search function" will comprehensively consider the relationship with these words and adjust its own vector representation to make it more accurately reflect its semantics in the context. Moreover, this information exchange is two-way, and "filter conditions" will also be affected by words such as "search function". This dynamic adjustment will continue in multiple Transformer layers, and the vector representation of each word will be continuously updated and refined. Through information transmission between layers, semantic information at different levels will be gradually integrated into the final vector representation. For longer demand texts, this feature of the Bert model is particularly important because it can well capture long-distance semantic dependencies. For example, in a long sentence, the first word can establish a semantic connection with the last word, overcoming the difficulties of traditional language models in dealing with long-distance dependencies.

[0025] After being processed by multiple layers of Transformer layers, the output vector is finally obtained. For the first requirement text, the semantic parsing encoding vector of the first requirement text is obtained, and for the second requirement text, the semantic parsing encoding vector of the second requirement text is obtained. Generally, it is necessary to pay attention to the output vector corresponding to the [CLS] tag at the beginning of each sentence, because it is considered to be the semantic representation of the entire sentence and can comprehensively reflect the semantic information of the entire sentence. Taking the first requirement text as an example, the output vector corresponding to the [CLS] tag will integrate the various words in the entire sentence and the relationship between them, and contain comprehensive semantic information on various functions and user experience requirements required for the development of an online shopping platform. At the same time, the average or maximum value of the output vectors of all words in the entire sentence can also be taken as the semantic representation of the sentence as needed, depending on the specific application scenario and different requirements for semantic information. These obtained semantic parsing encoding vectors contain rich semantic information of the input requirement text, not only the simple meaning of the word, but also the relationship between words and the overall semantics of the sentence.

[0026] In an embodiment of the present application, the business rule semantic parsing module 130 is used to extract business rules and perform semantic parsing on the business rules to obtain business rule semantic parsing encoding vectors. Specifically, in an embodiment of the present application, the business rule semantic parsing module is used to: use the semantic encoder containing the Bert model to perform semantic parsing on the business rules to obtain the business rule semantic parsing encoding vector. It should be understood that business rules contain information related to business strategies and goals, business process specification information, business resource allocation information, etc. In detail, business strategies and goals provide a macro direction for demand priority setting. Demands that meet long-term strategies or can help achieve short-term goals are usually given a higher priority because they directly serve the core business demands of the enterprise and help the enterprise gain competitive advantages or achieve phased results in the market. Business process specifications determine the closeness between demand and business operation. The priority of requirements related to key process steps and time constraints will be adjusted due to the impact on the smoothness and timeliness of business processes. The allocation of business resources limits the possibility and feasibility of demand realization. Demands with sufficient manpower and financial support will have more advantages in priority determination due to better implementation conditions. Considering that the business rules of an enterprise are formulated based on business goals, extracting business rules can make the determination of demand priorities closely revolve around business goals. For example, if the business goal of an enterprise is to improve user satisfaction, extracting related business rules, such as "improving customer service response speed, 90% of inquiries need to be replied within 10 minutes", can make it clear that demands such as "optimizing the intelligent response function of the customer service system" should have a higher priority, ensuring the consistency of demands with business goals, and avoiding waste of resources on demands that are not related to business goals. In addition, business rules are not static and will change with factors such as market environment, industry competition, and corporate strategy adjustments. Extracting business rules can perceive these changes in a timely manner and adjust demand priorities accordingly. Next, considering that business rules often have complex semantic connotations, it is difficult to deeply explore their internal logic in a simple text form, and there are semantic relationships between the various parts of the business rules. Therefore, in order to capture the deep semantic information in these rules and ensure a more accurate and comprehensive understanding of the rules, this application obtains business rule semantic parsing encoding vectors by semantically parsing the business rules. In particular, in a specific embodiment of the present application, the semantic encoder including the Bert model is used to perform semantic analysis on the business rules to utilize the powerful semantic understanding ability of the Bert model to deeply analyze complex business rules, sort out the logical relationship between different conditions of the business rules, accurately grasp the connotation of the rules, and obtain the semantic analysis encoding vector of the business rules, so as to comprehensively reflect the true meaning of the rules and their internal logic, and lay a solid foundation for further intelligent analysis and decision-making.

[0027] In an embodiment of the present application, the demand difference calculation module 140 is used to calculate the demand difference semantic parsing coding vector between the first demand text semantic parsing coding vector and the second demand text semantic parsing coding vector. Specifically, in an embodiment of the present application, the demand difference calculation module is used to: calculate the positional difference between the first demand text semantic parsing coding vector and the second demand text semantic parsing coding vector to obtain the demand difference semantic parsing coding vector. It should be understood that the first demand text semantic parsing coding vector and the second demand text semantic parsing coding vector respectively represent the representations of two different demand texts after semantic quantization, in which there are different difference information. Therefore, in order to capture and refine the subtle differences between the two at the semantic level, the present application calculates the positional difference between the first demand text semantic parsing coding vector and the second demand text semantic parsing coding vector to clearly show the differences between the two demand texts in various semantic details, obtain the demand difference semantic parsing coding vector, and provide data support for subsequent priority quantization. In the process of demand priority determination, this quantified difference information is an important reference basis. For example, if two requirement texts differ greatly in key semantic dimensions, it means that they have obvious differences in functions, goals, etc. According to business rules, different priorities may need to be set to help the system judge the priority of requirements more objectively and accurately.

[0028] In the embodiment of the present application, the demand difference bidirectional evaluation module 150 is used to perform a demand difference bidirectional attention evaluation based on business rules on the demand difference semantic parsing coding vector and the business rule semantic parsing coding vector to obtain a demand difference evaluation semantic coding vector under the business rule. Specifically, Figure 3 FIG. 1 is a block diagram of a demand difference bidirectional evaluation module in a demand priority intelligent assignment system based on demand semantic analysis according to an embodiment of the present application. Figure 3As shown, the demand difference bidirectional evaluation module 150 includes: a projection transformation encoding unit 151, which is used to perform homography projection transformation on the demand difference semantic parsing encoding vector and the business rule semantic parsing encoding vector to obtain a demand difference semantic parsing homography projection encoding vector and a business rule semantic parsing homography projection encoding vector; a forward and reverse semantic score field calculation unit 152, which is used to obtain a forward demand difference-business rule semantic parsing attention score field and a reverse demand difference-business rule semantic parsing attention score field based on the demand difference semantic parsing homography projection encoding vector and the business rule semantic parsing homography projection encoding vector. A scoring field; a bidirectional attention balance field construction unit 153, used to construct a positive and negative demand difference-business rule semantic analysis bidirectional attention balance field between the positive demand difference-business rule semantic analysis attention score field and the reverse demand difference-business rule semantic analysis attention score field; a demand difference-business rule semantic modulation response unit 154, used to modulate and interact the demand difference semantic analysis homography projection coding vector and the business rule semantic analysis homography projection coding vector based on the positive and negative demand difference-business rule semantic analysis bidirectional attention balance field to obtain the demand difference evaluation semantic coding vector under the business rule.

[0029] It should be understood that business rules are the criteria for business operations, and demand difference evaluation must be conducted within the framework of business rules to be meaningful. For example, in e-commerce business, business rules may stipulate that "demands directly related to improving user purchase conversion rates shall be given priority". Therefore, in order to be able to make more accurate priority assignments based on business rules, in the technical solution of the present application, a bidirectional attention evaluation mechanism for demand differences based on business rules is introduced to process the demand difference semantic parsing coding vector and the business rule semantic parsing coding vector to obtain a demand difference evaluation semantic coding vector under business rules. In particular, this mechanism can effectively capture the complex semantic associations between the two. For example, demand differences may be reflected in multiple aspects, and different clauses in business rules may have different considerations for these differences. This can help the system pay attention to the parts of the demand difference vector related to the clauses of the business rules and the parts of the business rule vector that are sensitive to demand differences during evaluation, so as to understand the relationship between them more comprehensively and accurately.

[0030] Specifically, the requirement difference semantic parsing coding vector and the business rule semantic parsing coding vector are firstly subjected to homography projection transformation to obtain the requirement difference semantic parsing homography projection coding vector and the business rule semantic parsing homography projection coding vector. The above process can be expressed as: ;in, is the semantic parsing encoding vector of the demand difference, is the business rule semantic parsing encoding vector, is the required difference mapping homography matrix, is the business rule mapping homography matrix, is the homography projection encoding vector of the demand difference semantic parsing, It is the homography projection encoding vector of business rule semantic parsing.

[0031] It should be understandable that the feature space formed by the original demand difference semantic analysis coding vector and the business rule semantic analysis coding vector may only reflect the surface or conventional feature relationship. In order to be able to dig out the potential relationship hidden behind the data and better reveal the deeper connection between demand and rules, this application needs to perform homography projection transformation on the demand difference semantic analysis coding vector and the business rule semantic analysis coding vector to map these two coding vectors to a new feature space. In the new feature space, the transformed demand difference semantic analysis homography projection coding vector and business rule semantic analysis homography projection coding vector can present information from different perspectives and scales, which helps the model to more accurately realize demand priority judgment.

[0032] Specifically, in the embodiment of the present application, the forward and reverse semantic score field calculation unit is used to calculate the forward demand difference-business rule semantic parsing attention score field of the demand difference semantic parsing homography projection coding vector relative to the business rule semantic parsing homography projection coding vector. The process can be expressed by the formula: ;in, is the homography projection encoding vector of the demand difference semantic parsing, represents matrix multiplication, yes The transposed vector of yes Length, It is the positive demand difference-business rule semantic parsing attention score field.

[0033] The reverse demand difference-business rule semantic parsing attention score field of the business rule semantic parsing homography projection coding vector relative to the demand difference semantic parsing homography projection coding vector is calculated. The process can be expressed by the formula: ;in, is the homography projection encoding vector of business rule semantic parsing, yes The transposed vector of is matrix multiplication, yes Length, It is the reverse demand difference-business rule semantic parsing attention score field.

[0034] It should be understood that in the process of determining demand priorities, it is necessary to clearly understand the close connection between demand differences and business rules. By calculating the positive demand difference-business rule semantic parsing attention score field, the importance distribution of the demand difference semantic parsing homography projection coding vector relative to the business rule semantic parsing homography projection coding vector can be quantified. For example, demand differences may be reflected in product display optimization, payment process simplification, etc. Calculating the positive demand difference-business rule semantic parsing attention score field can clarify the importance of these demand differences in the context of the business rule of improving purchase conversion rate, thereby providing a key basis for demand priority judgment. Moreover, in actual business, demand differences often contain a lot of information, and not all parts are closely related to business rules. The calculation of the positive demand difference-business rule semantic parsing attention score field is similar to the focusing mechanism of the human visual system, which enables the model to focus on the part of the demand difference that best reflects the relevance to the business rules. For example, in a software development project, demand differences may involve multiple ways of implementing functions, different user interface designs, and other aspects, but business rules emphasize the stability and compatibility of the system. By calculating the attention score field, the model can focus on the parts of the requirement differences that are related to stability and compatibility, and ignore those information that have little impact on business rules, thereby more accurately evaluating the degree to which the requirements meet the business rules.

[0035] Accordingly, considering that only relying on the forward demand difference-business rule semantic parsing attention score field may not cover all the associated information between demand differences and business rules. The present application can examine the relationship between the two from the opposite direction and supplement the details that may be missed in the forward analysis by calculating the reverse demand difference-business rule semantic parsing attention score field of the business rule semantic parsing homography projection coding vector relative to the demand difference semantic parsing homography projection coding vector. In addition, in actual business scenarios, the relationship between demand differences and business rules is often not symmetrical. The forward attention score field reflects the importance of demand differences in the context of business rules, while the reverse score field can reflect the unique impact of business rules in the context of demand differences. By calculating the reverse demand difference-business rule semantic parsing attention score field, this asymmetric relationship can be captured, which in turn helps to improve the accuracy of subsequent demand priority determination.

[0036] Then, a positive and negative demand difference-business rule semantic analysis bidirectional attention balance field is constructed between the positive demand difference-business rule semantic analysis attention score field and the reverse demand difference-business rule semantic analysis attention score field. The above process can be expressed as: ;in, It is the positive demand difference-business rule semantic analysis attention score field, It is the reverse demand difference-business rule semantic analysis attention score field, is the feature concatenation operation, is a convolutional code with a convolution kernel of 3×3. It is the difference between positive and negative demands - business rules semantic analysis two-way attention balance field.

[0037] It should be understood that the separate positive demand difference-business rule semantic parsing attention score field and the reverse demand difference-business rule semantic parsing attention score field provide information about the relationship between demand differences and business rules from different directions. However, in order to more comprehensively and accurately evaluate the interaction between the two, it is necessary to integrate the information from these two aspects. Based on this, in this application, it is necessary to construct a positive and negative demand difference-business rule semantic parsing bidirectional attention balance field between the positive demand difference-business rule semantic parsing attention score field and the reverse demand difference-business rule semantic parsing attention score field. The constructed positive and negative demand difference-business rule semantic parsing bidirectional attention balance field ensures that in the final interactive response encoding, the contributions of the positive and reverse scoring fields are reasonably considered, and the importance of one scoring field will not be ignored due to the excessive influence of the other.

[0038] Specifically, in an embodiment of the present application, the demand difference-business rule semantic modulation response unit is used to: map the demand difference semantic parsing homography projection coding vector and the business rule semantic parsing homography projection coding vector to the positive and negative demand difference-business rule semantic parsing bidirectional attention balance field to obtain the demand difference semantic parsing homography projection attention modulation coding vector and the business rule semantic parsing homography projection attention modulation coding vector, and the process can be expressed by the formula: ;in, is the homography projection encoding vector of the demand difference semantic parsing, is the homography projection encoding vector of business rule semantic parsing, It is the difference between positive and negative demands - business rules semantic analysis two-way attention balance field, is matrix multiplication, is the demand difference semantic parsing homography projection attention modulation encoding vector, It is the business rule semantic parsing homography projection attention modulation encoding vector.

[0039] The semantic coding vector of the demand difference evaluation under the business rule is obtained by dividing the position point by the homography projection attention modulation coding vector of the semantic analysis of the demand difference and the homography projection attention modulation coding vector of the semantic analysis of the business rule. The process can be expressed by the formula: ;in, is the demand difference semantic parsing homography projection attention modulation encoding vector, is the business rule semantic parsing homography projection attention modulation encoding vector, It is the semantic encoding vector of the demand difference evaluation under the business rules.

[0040] It should be understood that mapping the homography projection coding vector of demand difference semantic parsing and the homography projection coding vector of business rule semantic parsing to the positive and negative demand difference-business rule semantic parsing bidirectional attention balance field realizes the deep fusion of the two vector information. Specifically, the original coding vectors carry the relevant information of demand difference and business rules respectively, while the balance field integrates the information of the mutual influence between the two. Through mapping, the newly generated demand difference semantic parsing homography projection attention modulation coding vector and business rule semantic parsing homography projection attention modulation coding vector not only retain their original characteristics, but also incorporate the influence of each other's characteristics. For example, in e-commerce business, the demand difference coding vector may involve multiple solutions for optimizing product display, and the business rule coding vector revolves around the rules for improving sales and user satisfaction. After mapping, the new demand difference modulation coding vector will incorporate the requirements of business rules for sales and user satisfaction, thereby more comprehensively reflecting the actual significance of the demand under the framework of business rules.

[0041] Accordingly, considering that in the process of determining demand priorities, it is not enough to only understand the respective characteristics of demand differences and business rules, it is more critical to clarify the relationship between them. Although the previous steps modulated the homography projection coding vector of semantic analysis of demand differences and the homography projection coding vector of semantic analysis of business rules to integrate the influence of each other, it is still necessary to further explore the new relationship generated by this fusion. As an effective comparison mechanism, the dot division operation can deeply analyze the two modulated feature vectors and reveal the hidden and deep connection between them. Based on this, in this application, it is necessary to calculate the position point division between the homography projection attention modulation coding vector of semantic analysis of demand differences and the homography projection attention modulation coding vector of semantic analysis of business rules to obtain the semantic coding vector of demand difference evaluation under the business rules. The semantic coding vector of demand difference evaluation under the business rules obtained by performing the dot division operation can reveal the subtle proportional relationship between the features, and can provide a detailed and critical basis for the subsequent determination of demand priorities.

[0042] In an embodiment of the present application, the assignment result generation module 160 is used to obtain an assignment result based on the semantic coding vector of the demand difference evaluation under the business rules. Specifically, in an embodiment of the present application, the assignment result generation module is used to: input the semantic coding vector of the demand difference evaluation under the business rules into the demand priority assignment module based on the classifier to obtain the assignment result, and the assignment result is used to indicate that the priority of the first demand is higher than the priority of the second demand, the priority of the first demand is equal to the priority of the second demand, or the priority of the first demand is lower than the priority of the second demand. That is, the semantic coding vector of the demand difference evaluation under the business rules obtained by bidirectional evaluation of the demand difference semantic parsing coding vector and the business rule semantic parsing coding vector is classified and processed, so as to intelligently assign the priority of the first demand and the second demand. In particular, the classifier is a mature technology in the field of machine learning, with powerful pattern recognition and classification capabilities. It can extract key features from complex coding vector data and classify according to the patterns learned by training. For example, classifiers such as support vector machines and random forests perform well in processing high-dimensional data classification tasks, can effectively process multi-dimensional semantic information in coding vectors, and accurately judge the relationship between demand priorities. In this way, based on the clear priority assignment results, the project team can allocate resources more reasonably, including manpower, material resources and time, etc. Investing resources first in high-priority requirements can ensure that key requirements are met in a timely manner and improve the overall value and benefits of the project. In particular, in a specific embodiment of the present application, the semantic coding vector of the demand difference evaluation under the business rules is input into the demand priority assignment module based on the classifier to obtain the assignment result, and the assignment result is used to indicate that the priority of the first demand is higher than the priority of the second demand, the priority of the first demand is equal to the priority of the second demand, or the priority of the first demand is lower than the priority of the second demand, including: using the fully connected layer of the classifier to fully connect the semantic coding vector of the demand difference evaluation under the business rules to obtain the semantic fully connected coding feature vector of the demand difference evaluation under the business rules; inputting the semantic fully connected coding feature vector of the demand difference evaluation under the business rules into the Softmax classification function of the classifier to obtain the probability value of the semantic coding vector of the demand difference evaluation under the business rules belonging to each classification label, and the classification label includes a label used to indicate that the priority of a demand is higher than the priority of a second demand, a label used to indicate that the priority of the first demand is equal to the priority of the second demand, and a label used to indicate that the priority of the first demand is lower than the priority of the second demand; the classification label corresponding to the largest of the probability values ​​is determined as the assignment result.

[0043] In one example, considering that the requirement difference semantic analysis coding vector represents the analysis semantic difference between the first requirement text and the second requirement text, and the business rule semantic analysis coding vector represents the semantic analysis coding feature of the business rule, when performing an interactive response based on the feature positive and negative attention fields, the difference in the attribute of the semantic feature population encoded by the different feature patterns under the source semantics will cause a fairness difference in the reinforcement level of the positive and negative attention fields, thereby affecting the interactive response inclusiveness of the feature distribution of the semantic coding vector for evaluating the requirement difference under the business rules, and reducing the accuracy of the assignment result obtained by the requirement priority assignment module based on the classifier as input.

[0044] Therefore, in this example, when the semantic coding vector of the demand difference evaluation under the business rule is input into the demand priority assignment module based on the classifier to obtain the assignment result, the semantic coding vector of the demand difference evaluation under the business rule is optimized. The specific optimization process is: each eigenvalue of the semantic coding vector of the demand difference evaluation under the business rule is probabilistically calculated, for example, The function is activated to obtain the semantic encoding probability vector of demand difference evaluation under business rules, for example, .

[0045] Determine the eigenvalue mean of the semantic encoding probability vector of the demand difference evaluation under the business rule and the eigenvalue variance , and the semantic encoding probability vector of the difference evaluation requirement under the business rule Perform point multiplication to obtain the semantic encoding probability boundary restriction vector of the demand difference evaluation under the first business rule and the probability boundary restriction vector of the semantic encoding of the demand difference evaluation under the second business rule ,in, Represents dot product.

[0046] Multiply the point-by-point absolute value vector of the semantic coding probability boundary limit vector of the demand difference evaluation under the first business rule and the semantic coding probability boundary limit vector of the demand difference evaluation under the second business rule by the logarithm vector of the semantic coding probability vector of the demand difference evaluation under the business rule with base 2 to obtain the semantic coding probability information transmission vector of the demand difference evaluation under the business rule ,in, Indicates point reduction.

[0047] The bit-by-bit inverse of the semantic coding probability boundary limit vector of the demand difference evaluation under the first business rule is multiplied by the semantic coding probability boundary limit vector of the demand difference evaluation under the second business rule, and then the square root of the quotient of the mean of the eigenvalue divided by the standard deviation of the eigenvalue is added to obtain the multi-granularity fair coordination vector of the semantic coding of the demand difference evaluation under the business rule ,in, represents the bit-by-bit reciprocal of the semantically encoded probability boundary limit vector of the demand difference evaluation under the first business rule, where the bit-by-bit reciprocal is the reciprocal of the eigenvalue at each position of the calculation vector, Indicates point addition.

[0048] The semantic coding probability information transmission vector of the demand difference evaluation under the business rules is interpolated with the multi-granularity fair coordination vector of the semantic coding of the demand difference evaluation under the business rules to obtain the optimized semantic coding vector of the demand difference evaluation under the business rules .

[0049] That is, taking into account the differences in dynamic interaction modes of the semantic coding vectors for demand difference evaluation under the business rules, in order to improve the dynamic response tolerance of the semantic coding vectors for demand difference evaluation under the condition of representation distribution discreteness, a bidirectional fairness indicator based on mixed probability boundary constraints is adopted as the interaction constraint target, and the group characteristic information transmission process of the semantic coding vectors for demand difference evaluation under the business rules is optimized and adjusted. By introducing a hierarchical fairness calibration factor based on a global statistical response mechanism, a robust distribution consistency representation system is constructed under a multi-granularity fairness collaboration architecture of the semantic coding vectors for demand difference evaluation under the business rules, so as to form a fairness collaboration framework with adaptive adjustment capabilities, effectively balance the multi-dimensional interaction relationship of the semantic coding vectors for demand difference evaluation under the business rules in the feature space, and improve the accuracy of the assignment results obtained by the demand priority assignment module based on the classifier as input.

[0050] In summary, the demand priority intelligent assignment system 100 based on demand semantic analysis based on the embodiment of the present application is explained, which uses natural language analysis and coding technology based on artificial intelligence to semantically parse the first demand text and the second demand text, and at the same time, extracts business rules and performs semantic analysis on them, and then calculates the demand difference semantic analysis features between the semantic analysis features of the first demand text and the semantic analysis features of the second demand text, so as to intelligently assign the priority of the first demand and the second demand based on the two-way attention demand difference semantic representation guided by business rules between the demand difference semantic analysis features and the business rule semantic analysis features. In this way, the accuracy and rationality of demand priority evaluation can be effectively improved.

Claims

1. A demand priority intelligent assignment system based on demand semantic analysis, characterized in that: include: A requirement text acquisition module is used to acquire a first requirement text and a second requirement text; a requirement text semantic parsing module is used to perform semantic parsing on the first requirement text and the second requirement text to obtain a first requirement text semantic parsing coding vector and a second requirement text semantic parsing coding vector; a business rule semantic parsing module is used to extract business rules and perform semantic parsing on the business rules to obtain a business rule semantic parsing coding vector; a requirement difference calculation module is used to calculate a requirement difference semantic parsing coding vector between the first requirement text semantic parsing coding vector and the second requirement text semantic parsing coding vector; a requirement difference bidirectional evaluation module is used to perform a requirement difference bidirectional attention evaluation based on business rules on the requirement difference semantic parsing coding vector and the business rule semantic parsing coding vector to obtain a requirement difference evaluation semantic coding vector under business rules; and an assignment result generation module is used to obtain an assignment result based on the requirement difference evaluation semantic coding vector under the business rules.

2. The demand priority intelligent assignment system based on demand semantic analysis according to claim 1 is characterized in that: The requirement text semantic parsing module is used to: use a semantic encoder including a Bert model to perform semantic parsing on the first requirement text and the second requirement text to obtain a semantic parsing encoding vector of the first requirement text and a semantic parsing encoding vector of the second requirement text.

3. The demand priority intelligent assignment system based on demand semantic analysis according to claim 2 is characterized in that: The business rule semantic parsing module is used to: use the semantic encoder containing the Bert model to perform semantic parsing on the business rule to obtain the business rule semantic parsing encoding vector.

4. The demand priority intelligent assignment system based on demand semantic analysis according to claim 3 is characterized in that: The requirement difference calculation module is used to calculate the positional difference between the first requirement text semantic parsing encoding vector and the second requirement text semantic parsing encoding vector to obtain the requirement difference semantic parsing encoding vector.

5. The demand priority intelligent assignment system based on demand semantic analysis according to claim 4 is characterized in that: The demand difference bidirectional evaluation module includes: a projection transformation encoding unit, which is used to perform homography projection transformation on the demand difference semantic analysis encoding vector and the business rule semantic analysis encoding vector to obtain a demand difference semantic analysis homography projection encoding vector and a business rule semantic analysis homography projection encoding vector; a forward and reverse semantic score field calculation unit, which is used to obtain a forward demand difference-business rule semantic analysis attention score field and a reverse demand difference-business rule semantic analysis attention score field based on the demand difference semantic analysis homography projection encoding vector and the business rule semantic analysis homography projection encoding vector; a bidirectional attention balance field construction unit, which is used to construct a positive and negative demand difference-business rule semantic analysis bidirectional attention balance field between the forward demand difference-business rule semantic analysis attention score field and the reverse demand difference-business rule semantic analysis attention score field; a demand difference-business rule semantic modulation response unit, which is used to modulate and interact the demand difference semantic analysis homography projection encoding vector and the business rule semantic analysis homography projection encoding vector based on the positive and negative demand difference-business rule semantic analysis bidirectional attention balance field to obtain a demand difference evaluation semantic encoding vector under the business rule.

6. The demand priority intelligent assignment system based on demand semantic analysis according to claim 5 is characterized in that: The forward and reverse semantic score field calculation unit is used to: calculate the forward demand difference-business rule semantic parsing attention score field of the demand difference semantic parsing homography projection coding vector relative to the business rule semantic parsing homography projection coding vector; calculate the reverse demand difference-business rule semantic parsing attention score field of the business rule semantic parsing homography projection coding vector relative to the demand difference semantic parsing homography projection coding vector.

7. The demand priority intelligent assignment system based on demand semantic analysis according to claim 6 is characterized in that: The demand difference-business rule semantic modulation response unit is used to: map the demand difference semantic parsing homography projection coding vector and the business rule semantic parsing homography projection coding vector to the positive and negative demand difference-business rule semantic parsing bidirectional attention balance field respectively to obtain the demand difference semantic parsing homography projection attention modulation coding vector and the business rule semantic parsing homography projection attention modulation coding vector; The demand difference semantic parsing homography projection attention modulation coding vector and the business rule semantic parsing homography projection attention modulation coding vector are calculated and divided by the position points to obtain the demand difference evaluation semantic coding vector under the business rule.

8. The demand priority intelligent assignment system based on demand semantic analysis according to claim 7 is characterized in that: The assignment result generation module is used to: input the semantic coding vector of the requirement difference evaluation under the business rules into the classifier-based requirement priority assignment module to obtain the assignment result, and the assignment result is used to indicate that the priority of the first requirement is higher than the priority of the second requirement, the priority of the first requirement is equal to the priority of the second requirement, or the priority of the first requirement is lower than the priority of the second requirement.

Images