Method and device for evaluating tea leaf quality

Abstract

This specification provides a method and apparatus for tea quality evaluation. The method includes: acquiring quality evaluation information of the tea to be evaluated across multiple quality dimensions, and quality standard information of a standard tea sample across multiple quality dimensions; determining each target difference item and corresponding difference information between the tea to be evaluated and the standard tea sample based on the quality evaluation information of multiple quality dimensions and the corresponding quality standard information; and analyzing each target difference item and corresponding difference information based on the difference dimension constraints of the tea to be evaluated to determine the quality evaluation result of the tea to be evaluated. The multiple quality dimensions and difference dimension constraints are configured based on the attributes of the tea to be evaluated. The difference dimension constraints are used to indicate the constraint relationship between the difference items, difference information, and the quality evaluation result, including single-dimensional constraints and / or total-dimensional constraints. This achieves objectivity and comprehensiveness in the evaluation process and improves the accuracy of tea quality evaluation.

Description

[0001] This application claims priority to Chinese Patent Application No. 202511661011.9, filed on November 13, 2025, entitled “An Evaluation Method for Flower Tea Raw Materials Based on a Digital Sensory Quantitative Evaluation System”, the entire contents of which are recorded in this application by reference. Technical Field

[0002] The embodiments in this specification relate to the field of tea evaluation technology, and in particular to a method and apparatus for evaluating tea quality. Background Technology

[0003] Tea evaluation is a core step in determining tea quality. This process relies on evaluators using their senses—sight, smell, and taste—to comprehensively analyze and judge multiple core attributes of tea, including aroma, flavor, appearance, and liquor color. Traditional tea evaluation methods typically depend on evaluators using these senses to comprehensively assess tea quality. While this method can determine tea quality, it demands a high level of professional experience from the evaluators and is easily influenced by environmental factors and personal subjective preferences, making it difficult to guarantee the stability and accuracy of the evaluation results. Furthermore, under the traditional manual evaluation model, a universally standardized quality evaluation framework that can flexibly adapt to the characteristics of different tea types has not been established. The evaluation dimensions, judgment scales, and rules for different tea types largely depend on manual experience, making it difficult to flexibly adapt and standardize them under a unified framework based on the quality characteristics of different tea types such as green tea and black tea. This further exacerbates the subjectivity and inconsistency of the evaluation results. Therefore, there is an urgent need to establish an objective and stable tea quality evaluation method to improve the accuracy and reliability of the evaluation results. Summary of the Invention

[0004] In view of this, embodiments of this specification provide a method for evaluating tea. One or more embodiments of this specification also relate to a tea evaluation apparatus, a computing device, a computer-readable storage medium, and a computer program product, to address the technical deficiencies existing in the prior art.

[0005] According to a first aspect of the embodiments of this specification, a tea evaluation method is provided, comprising: Obtain quality assessment information of the tea to be evaluated in multiple quality dimensions, as well as quality standard information of the standard tea in multiple quality dimensions. The multiple quality dimensions are configured based on the attributes of the tea to be evaluated, and the difference dimension constraints include single dimension constraints and / or total dimension constraints. Based on quality assessment information from multiple quality dimensions and corresponding quality standard sample information, at least one target difference item between the tea to be evaluated and the standard tea in multiple quality dimensions is determined, along with the difference information corresponding to each target difference item. Based on the difference dimension constraints corresponding to the tea to be evaluated, at least one target difference item and the difference information corresponding to each target difference item are analyzed to determine the quality evaluation result of the tea to be evaluated. The difference dimension constraints are configured based on the attributes of the tea to be evaluated and are used to indicate the constraint relationship between the difference item, the difference information and the quality evaluation result. The difference dimension constraints include single dimension constraints and / or total dimension constraints.

[0006] According to a second aspect of the embodiments of this specification, a tea quality evaluation device is provided, comprising: The acquisition module is configured to acquire quality assessment information of the tea to be evaluated in multiple quality dimensions, as well as quality standard information of the standard tea in the multiple quality dimensions, wherein the multiple quality dimensions are configured based on the attributes of the tea to be evaluated. The first determining module is configured to determine at least one target difference item between the tea to be evaluated and the standard tea in multiple quality dimensions, and the difference information corresponding to each target difference item, based on quality assessment information of multiple quality dimensions and corresponding quality standard sample information. The second determining module is configured to analyze at least one target difference item and the difference information corresponding to each target difference item based on the difference dimension constraints corresponding to the tea to be evaluated, and determine the quality evaluation result of the tea to be evaluated. The difference dimension constraints are configured based on the attributes of the tea to be evaluated and are used to indicate the constraint relationship between the difference item, the difference information and the quality evaluation result. The difference dimension constraints include single dimension constraints and / or total dimension constraints.

[0007] According to a third aspect of the embodiments of this specification, a computing device is provided, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the above-described tea quality evaluation method.

[0008] According to a fourth aspect of the embodiments of this specification, a computer-readable storage medium is provided that stores computer-executable instructions, which, when executed by a processor, implement the steps of the tea quality evaluation method described above.

[0009] According to a fifth aspect of the embodiments of this specification, a computer program product is provided, including a computer program / instructions that, when executed by a processor, implement the steps of the above-described tea quality evaluation method.

[0010] One embodiment of this specification implements a tea evaluation method, comprising: acquiring quality evaluation information of tea to be evaluated in multiple quality dimensions, and quality standard information of standard tea in multiple quality dimensions, wherein the multiple quality dimensions are configured based on the attributes of the tea to be evaluated; determining at least one target difference item between the tea to be evaluated and the standard tea in multiple quality dimensions, and the difference information corresponding to each target difference item, based on the quality evaluation information of the multiple quality dimensions and the corresponding quality standard information; and analyzing at least one target difference item and the difference information corresponding to each target difference item, based on the difference dimension constraints corresponding to the tea to be evaluated, to determine the quality evaluation result of the tea to be evaluated, wherein the difference dimension constraints are configured based on the attributes of the tea to be evaluated and are used to indicate the constraint relationship between the difference item, the difference information and the quality evaluation result, and the difference dimension constraints include single dimension constraints and / or total dimension constraints. Thus, by accurately acquiring evaluation information and standard sample information across multiple quality dimensions of the tea to be evaluated and the standard sample, and clarifying the target differences and corresponding differences between the two, a refined quantitative characterization of tea quality differences is achieved, effectively overcoming the shortcomings of traditional sensory evaluation that relies on subjective experience and has vague difference descriptions. Simultaneously, by configuring differentiated difference dimension constraints based on the inherent attributes of the tea to be evaluated, the quality evaluation process is highly aligned with the core quality requirements of tea, improving the accuracy of tea quality evaluation. Furthermore, this method, through its core design of "attributed quality dimensions + standardized difference analysis process + attributed constraint configuration," successfully constructs a system that can flexibly adapt to the characteristics of different tea types. The universal standardized quality assessment framework—a standardized difference analysis process (acquiring assessment information → identifying target difference items → analyzing difference information)—provides a unified assessment benchmark for all tea types, breaking the predicament of traditional manual evaluation where different tea types lack unified standards and assessment logic is chaotic, thus achieving the universality of the assessment framework. Furthermore, by configuring quality dimensions based on the inherent attributes of the tea being evaluated, and designing differentiated constraints, the framework can flexibly adjust quality dimensions, difference judgment standards, constraint thresholds, and assessment focuses according to the core quality orientations and characteristics of different tea types such as green tea and black tea. This achieves precise adaptation to the characteristics of different tea types, enabling the same assessment framework to cover quality assessment scenarios for multiple tea types. Replacing traditional subjective manual assessment with standardized difference analysis and attribute-based constraint judgment significantly improves assessment efficiency, enabling the rapid output of clear and traceable quality assessment results. It can efficiently adapt to the quality control needs of multiple scenarios in industrial production, such as raw material batch acceptance, production process quality monitoring, and finished product inspection, providing solid technical support for the standardized, large-scale, and refined development of the tea industry. Attached Figure Description

[0011] Figure 1 This is a flowchart of a tea evaluation method provided in one embodiment of this specification; Figure 2 This is a schematic diagram of the structure of a tea full-chain quality control and supply chain optimization system provided in one embodiment of this specification; Figure 3 This is a flowchart illustrating the processing steps of a tea evaluation method provided in one embodiment of this specification. Figure 4 This is a schematic diagram of the structure of a tea evaluation device provided in one embodiment of this specification; Figure 5 This is a structural block diagram of a computing device provided in one embodiment of this specification. Detailed Implementation

[0012] Many specific details are set forth in the following description to provide a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this specification. Therefore, this specification is not limited to the specific implementations disclosed below.

[0013] The terminology used in one or more embodiments of this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of this specification. The singular forms “a” and “the” as used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0014] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this specification, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this specification, and similarly, second may also be referred to as first. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to a determination."

[0015] Furthermore, it should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in one or more embodiments of this specification are all information and data authorized by the user or fully authorized by all parties. Moreover, the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0016] This specification provides a method for evaluating tea quality. It also relates to a tea quality evaluation device, a computing device, a computer-readable storage medium, and a computer program product, which will be described in detail in the following embodiments.

[0017] See Figure 1 , Figure 1 A flowchart of a tea quality evaluation method according to an embodiment of this specification is shown, specifically including steps 102 to 106.

[0018] Step 102: Obtain quality assessment information of the tea to be evaluated in multiple quality dimensions, and quality standard information of the standard tea in multiple quality dimensions. The multiple quality dimensions are configured based on the attributes of the tea to be evaluated.

[0019] Among them, tea samples awaiting evaluation refer to tea samples whose quality grade is currently unknown and needs to be evaluated. These usually come from newly arrived production batches, raw material samples provided by suppliers, or finished products awaiting inspection on the production line.

[0020] Quality dimensions refer to a structured sensory evaluation index system for tea. This system adopts a flexibly configurable hierarchical structure, supporting the independent use of basic dimensions (such as appearance, liquor color, aroma, taste, and infused leaf appearance) and allowing for the extension of one or more sub-dimensions according to the needs of specific quality control scenarios. The dimensional structure can adopt a flat mode (no sub-dimensions) or a tree-like hierarchical mode (containing multiple sub-dimensions) depending on the required level of granularity in the evaluation, thereby achieving multi-scale measurement and standardized control of tea quality characteristics.

[0021] For example, the quality dimensions include five main dimensions: appearance, liquor color, aroma, taste, and infused leaf residue. Each dimension can be further subdivided into several quantifiable evaluation sub-dimensions, as detailed below: Appearance: further subdivided into shape, strands, tenderness, color, and uniformity; Soup color: further subdivided into color and brightness; Aroma: Subdivided into characteristic aroma (including aroma type, purity, intensity, and persistence) and unusual odor (including type and intensity of unusual odor). Taste: Subdivided into characteristic taste (including flavor profile, aroma, and consistency), body (including concentration and thickness), bitterness (including bitterness, astringency, and the speed at which astringency dissipates), and unusual taste (including off-flavor type and intensity). Leaf base: further subdivided into tenderness, color (including color and brightness), and uniformity.

[0022] Furthermore, the quality dimensions in the embodiments of this specification can be dynamically set according to the characteristics of the evaluation object itself.

[0023] The attributes of the tea leaves to be evaluated in the embodiments of this specification refer to the key characteristics used to distinguish the type, grade, and use of tea leaves, and may include, but are not limited to: Tea categories: green tea, flower tea, black tea, oolong tea, etc. Level attributes: High-end / Common levels such as Special Grade, Grade 1, Grade 2, etc. Product attributes: raw tea, semi-finished tea, finished tea, etc.

[0024] In the embodiments of this specification, based on the above-mentioned attributes of the tea to be evaluated, quality dimensions will be selected, added, deleted, or adjusted in a targeted manner: different main dimensions can be configured to adapt to the core evaluation direction of different tea types; different sub-dimensions can also be configured to refine the evaluation granularity of the corresponding main dimensions; ultimately forming a unique quality dimension system that matches the characteristics of the tea itself.

[0025] For example, for green tea, appearance, aroma, taste, liquor color, and infused leaf are set as main dimensions, and sub-dimensions such as leaf shape, color, and integrity are set under the main dimension of appearance. Under the main dimension of aroma, sub-dimensions such as freshness and crispness are set. For flower tea with prominent floral characteristics, the main dimensions can be adjusted and added based on its core quality attributes. Aroma quality and taste harmony are set as separate main dimensions, while the main dimensions of appearance, liquor color, and infused leaf are retained. Under the main dimension of aroma quality, sub-dimensions such as aroma purity, floral fragrance layering, and aroma persistence are set. For high-grade tea, the number of main dimensions can be refined and the sub-dimensions can be enriched. For ordinary-grade tea, the main and sub-dimension settings can be simplified, thereby achieving flexible adaptation of quality evaluation dimensions for different tea types and grades.

[0026] Quality assessment information refers to the raw data or standardized descriptions obtained by evaluators or sensors after measuring or sensing the tea leaves to be evaluated for each of the above quality dimensions. For example, a score of "8" is given for the "aroma intensity" dimension. Another example is selecting one from a predefined set of descriptions for each quality dimension, such as checking "rich" or "refined" for the "aroma intensity" dimension.

[0027] Standard tea samples are pre-determined tea samples that represent the ideal quality state of a certain type of tea. Standard tea samples are usually "benchmark samples" determined by a team of experts at the beginning of the production season.

[0028] Quality standard sample information refers to the ideal values ​​or descriptions of a standard tea sample across all quality dimensions, as certified by authoritative bodies. It is a structured database or form.

[0029] For example: Category: Premium Jasmine Tea, its standard sample is: Aroma - Intensity: 9 points (rich); Taste - Bitterness: 2 points (weak); Liquor Color - Brightness: 8 points (bright).

[0030] It should be noted that the embodiments in this specification do not limit the specific implementation method of obtaining quality assessment information of the tea to be evaluated in multiple quality dimensions.

[0031] In one optional implementation of this embodiment, the quality assessment information of the tea to be evaluated across multiple quality dimensions is obtained through the following method: The system obtains the voice descriptions of the reviewers for each quality dimension. After recognizing the voice, the system matches it with the preset tea quality standard terminology library and score mapping library, and automatically converts it into standardized quality assessment information.

[0032] Specifically, the system pre-constructs a standard terminology library (such as aroma: fresh, rich, light, and mixed; taste: mellow, bitter, and quick to dissipate) and score mapping relationship bound to tea attributes; Reviewers describe the target quality dimensions via voice input (e.g., "The sample's soup color is not bright enough, and the aroma is not fresh enough"). The system performs text recognition and semantic parsing on speech, and forces natural language to be mapped into standard terms / standard scores; The system automatically aggregates standardized data mapped from each dimension to form quality assessment information in a unified format.

[0033] To further ensure the standardization of data, non-standard descriptions will be automatically corrected / rejected by the system, and only standard terms and scores within the preset range will be output.

[0034] In another optional implementation of this embodiment, the quality assessment information of the tea to be evaluated across multiple quality dimensions is obtained through the following method: The system preloads standardized scoring templates for the corresponding quality dimensions based on the attributes (category, grade, and raw material type) of the tea to be evaluated. Reviewers complete the evaluation by selecting template items without the need for custom input.

[0035] Specifically, based on the attributes of the tea to be evaluated, the system automatically calls the matching quality dimension scoring template (e.g., a template specifically for jasmine green tea or black tea). The template pre-defines the standard score ranges and standard descriptions for each dimension; reviewers only need to select the corresponding level / score within the template. The system directly reads the template selection results and generates structured and standardized quality assessment information.

[0036] In another optional implementation of this embodiment, the quality assessment information of the tea to be evaluated across multiple quality dimensions is obtained through the following method: The system directly collects physical or chemical signals of tea leaves using sensory acquisition devices, and automatically generates standardized quality assessment information according to preset rules, eliminating the need for manual or voice input by reviewers.

[0037] For example, the following sensor devices can be connected: Appearance / Soup Color: High-definition image acquisition equipment → identifies color, brightness, and strand shape; Aroma: Electronic nose / odor sensor → identifies characteristic aromas, off-odors, and aroma intensity; Taste: Electronic tongue / physicochemical testing equipment → identifies bitterness, intensity, and flavor profile.

[0038] Specifically, the data acquisition equipment collects objective data on various quality dimensions of the tea to be evaluated. The system converts the objective signals collected by the equipment into standard scores / standard labels according to a preset model; and directly outputs the quality assessment information of the tea to be evaluated.

[0039] In another optional implementation of this embodiment, the quality assessment information of the tea to be evaluated in multiple quality dimensions is obtained, specifically including: The tea evaluation interface displays evaluation controls for multiple quality dimensions. These controls include preset quality information and / or input controls. The preset quality information is configured based on the attributes of the tea to be evaluated. In response to the evaluation operation of the evaluation control under the target quality dimension, the quality evaluation information under the target quality dimension is determined based on the preset quality information selected by the evaluation operation and / or the input quality information, wherein the target quality dimension is any one of multiple quality dimensions; In response to the completion event in the tea evaluation interface, obtain quality evaluation information for each quality dimension.

[0040] The tea evaluation interface is a specially designed software user interface (UI), which can be an application screen on a computer, tablet, or mobile phone. Its purpose is to guide users in completing the tea evaluation process.

[0041] Evaluation controls are UI elements displayed on the interface for user interaction. Each quality dimension has a corresponding control. They serve as a "tool" for users to input evaluation information. Types of evaluation controls include: Controls used to select "preset quality information" typically include radio button groups, drop-down lists, and rating sliders. For example, under the "aroma intensity" dimension, radio buttons such as "rich," "strong," "light," and "neutral" are provided for selection.

[0042] "Input controls": These are typically text boxes. They allow evaluators to enter custom descriptive text or ratings.

[0043] Preset quality information: This is a set of standardized descriptive options and / or scores prepared in advance by the system for each dimension. It is not fixed, but intelligently retrieved based on the "attributes of the tea to be evaluated". For example, when evaluating green tea, the preset options under the "aroma" dimension might be "fresh aroma, bean aroma, chestnut aroma"; while when evaluating black tea, the preset options would become "sweet aroma, potato aroma, fruit aroma".

[0044] Among them, tea attributes refer to the core metadata used to define and distinguish different tea sample categories and flavor characteristics. These typically include type (e.g., green tea, black tea, oolong tea), origin (e.g., West Lake, Wuyi Mountain, Menghai, Yunnan), grade (e.g., premium grade, first grade), season (e.g., spring tea, autumn tea), and processing characteristics (e.g., light roasting, heavy fermentation). By identifying these attributes, the system dynamically calls upon appropriate evaluation dimensions, terminology sets, and judgment rules to achieve precise and customized quality assessment.

[0045] Evaluation actions refer to the interactive actions performed by the user on this interface. Examples include: clicking a radio button, selecting an option from a drop-down list, entering text in a text box, and dragging a rating slider.

[0046] Because the interface displays many dimensions, and users evaluate each dimension one by one, the target quality dimension refers to the specific dimension that the user is currently operating on.

[0047] The evaluation completion event is a trigger signal indicating that the user has completed the evaluation work across all dimensions. It is typically triggered by the user clicking a "Submit" or "Complete" button on the interface.

[0048] Once the system "responds" to this event, it will begin executing subsequent processes, such as packaging all the evaluation information for all dimensions, sending it to the server, and initiating automatic calculation and judgment logic.

[0049] In the embodiments described in this specification, an intelligent interactive interface transforms the original subjective sensory evaluation process, which relied on personal experience, into a structured and standardized data collection process. This not only greatly improves efficiency, but more importantly, it ensures the standardization and consistency of data from the source, which is a crucial prerequisite for the success of the entire automated evaluation system.

[0050] Step 104: Based on the quality assessment information of multiple quality dimensions and the corresponding quality standard sample information, determine at least one target difference item between the tea to be evaluated and the standard tea in multiple quality dimensions, as well as the difference information corresponding to each target difference item.

[0051] The system automatically identifies differences between the tea being evaluated and standard tea by comparing their quality data dimension by dimension. Specifically, the system first determines whether there is a difference between the two sets of data on the same dimension, and then determines whether the difference belongs to the "valid difference" that needs attention based on preset difference judgment rules (such as difference threshold, terminology change, etc.). If it is judged to be a valid difference, the dimension is marked as the target difference item.

[0052] Difference information is used to quantitatively or qualitatively describe specific differences in target difference items, and can be divided into two forms of representation: Numerical differences: For example, the "aroma intensity" dimension shows "-1", which means that the intensity score of the tea to be evaluated is 1 point lower than that of the standard sample; Descriptive differences: For example, in the dimension of "bitterness and astringency", the standard sample is "slightly astringent" while the tea to be evaluated is "none", directly describing the change in this term.

[0053] For example, using jasmine green tea as the tea to be evaluated, the system pre-stores the quality standard sample information of jasmine green tea in each quality dimension, and simultaneously obtains the quality evaluation information of the tea to be evaluated in the same quality dimension. The preset business rule in this example is: under the same quality dimension, if the absolute value of the difference between the score of the tea to be evaluated and the score of the standard tea is ≥1, it is determined that there is a valid difference in that dimension, and that dimension is marked as the target difference item; if the absolute value of the difference is <1, it is determined that there is no valid difference.

[0054] Quality standard sample information (standard sample scoring): Appearance: 12.0 points; Broth color: 4.0 points; Aroma: 35.0 points; Taste: 31.0 points; Leaf appearance: 8.0 points; Quality assessment information for tea leaves to be evaluated (scores for samples to be tested): Appearance: 11.2 points; Broth color: 4.0 points; Aroma: 32.2 points; Taste: 27.9 points; Leaf base: 6.0 points; The system compares and identifies valid differences dimension by dimension: External dimensions: |11.2-12.0|=0.8<1, no valid difference, not marked as a target difference item; Soup color dimension: |4.0-4.0|=0<1, no valid difference, not marked as a target difference item; Aroma dimension: |32.2-35.0|=2.8≥1, there is a valid difference, marked as the target difference item; Taste dimension: |27.9-31.0|=3.1≥1, there is a valid difference, marked as the target difference item; Leaf base dimension: |6.0-8.0|=2.0≥1, there is a valid difference, marked as the target difference item.

[0055] System output results: The target differences between the tea to be evaluated and the standard tea are: aroma dimension, taste dimension, and leaf residue dimension; the difference information corresponding to each target difference is as follows: The aroma score was 2.8 points lower than the standard sample; The taste score was 3.1 points lower than the standard sample; The leaf base dimension score was 2.0 points lower than that of the standard sample.

[0056] In one optional implementation of this embodiment, the quality assessment information includes quality descriptive terms and / or quality scores, and the quality standard sample information includes standard sample descriptive terms and / or standard sample scores; based on the quality assessment information of multiple quality dimensions and the corresponding quality standard sample information, the target differences between the tea to be evaluated and the standard tea in multiple quality dimensions and the difference information corresponding to each target difference item are determined, including: By comparing the quality descriptors and / or quality scores under multiple quality dimensions, as well as the standard sample descriptors and / or standard sample scores, the target differences between the tea to be evaluated and the standard tea in multiple quality dimensions, and the difference information corresponding to each target difference, are determined.

[0057] This includes the evaluators' assessment results of the tea being evaluated. These results can take the following form: Quality description words: words selected from a preset terminology library (e.g., aroma type: floral, taste: full-bodied).

[0058] Quality rating: The score given directly (e.g., aroma intensity: 8 points).

[0059] Or both: for example, choosing "fragrant flowers" (descriptive word) and also giving "8 points" (rating).

[0060] The information on quality standards is in the same format as described above, and also includes standard descriptions and / or standard scores.

[0061] In this embodiment, the quality assessment information is the information obtained by the evaluator or the system after evaluating the tea to be evaluated, which may include quality descriptive words (such as fresh, rich, bright, etc.) and / or quality scores (quantified scores); the quality standard sample information is the standard sample benchmark information preset by the system, which includes standard sample descriptive words and / or standard sample scores, and serves as a reference benchmark for quality comparison.

[0062] During the difference identification process, the system performs a horizontal comparison across multiple quality dimensions, including appearance, liquor color, aroma, taste, and infused leaf appearance. It compares the quality descriptions and / or quality scores of the tea to be evaluated in each dimension with those of a standard tea sample in the same dimension. Based on preset difference judgment rules, it determines whether a valid difference exists in that dimension. If a dimension is determined to have a valid difference, that dimension is identified as the target difference item, and the corresponding difference information (such as score difference, description difference, and high / low deviation) is recorded and generated simultaneously.

[0063] In the embodiments of this specification, by adopting a flexible configuration of "quality descriptive terms and / or quality scores," quality assessment information and quality standard sample information can exist and be compared separately as quality descriptive terms, separately as quality scores, or a combination of both. This greatly improves the applicability and flexibility of the solution, and can perfectly adapt to different tea sensory evaluation scenarios such as qualitative description only, quantitative scoring only, and a combination of qualitative and quantitative methods. At the same time, the system can perform a dimension-by-dimensional standardized comparison of any of the above forms of information, automatically identifying the target differences and corresponding difference information between the tea to be evaluated and the standard tea. This ensures semantic consistency at the qualitative description level and achieves precise numerical comparison at the quantitative scoring level, effectively avoiding the subjectivity, ambiguity, and arbitrariness brought about by manual comparison, and greatly improving the standardization, objectivity, and stability of difference identification.

[0064] In an optional implementation of this embodiment, the quality assessment information includes quality descriptive terms, and the quality standard sample information includes standard sample descriptive terms. The quality descriptive terms across multiple quality dimensions and the standard sample descriptive terms are compared to determine the target differences between the tea to be evaluated and the standard tea across multiple quality dimensions, as well as the difference information corresponding to each target difference, including: The quality descriptors under multiple quality dimensions are compared with the standard descriptors to determine the degree of matching between the quality descriptors under multiple quality dimensions and the standard descriptors. If the degree of matching between the quality descriptor of the first quality dimension and the standard descriptor is less than the first threshold, then the first quality dimension is determined as the target difference item, wherein the first quality dimension is any one of multiple quality dimensions. The degree of matching between the quality descriptors of the first quality dimension and the standard sample descriptors is determined as the difference information corresponding to the target difference item.

[0065] The embodiments in this specification do not limit the specific method for determining the degree of matching between quality descriptors and standard sample descriptors under each quality dimension.

[0066] In one implementation, the degree of consistency between the two terms in terms of word meaning, sensory features, and descriptive level can be calculated based on preset semantic matching rules, and this degree of consistency can be quantified into a matching degree value (e.g., 0-100% or 0-1). The preset semantic matching rules include: if the quality descriptor and the sample descriptor are exactly the same, the matching degree is the highest value; if they are synonyms or similar descriptions, the matching degree is a higher intermediate value; if they are different descriptions, opposite descriptions, or have significant sensory differences, the matching degree is a lower value.

[0067] In one implementation, the degree of matching between quality descriptors and standard descriptors under each quality dimension is determined. Specifically, a semantic similarity model can be used for calculation: the quality descriptors and standard descriptors under the same quality dimension are respectively input into a pre-trained or built-in text semantic similarity model. The model extracts and compares the semantic features of the two descriptive texts, calculates the semantic similarity score between them, and directly uses the semantic similarity score as the degree of matching between the quality descriptors and standard descriptors under that quality dimension.

[0068] Among them, the higher the semantic similarity score, the closer the quality descriptor and the standard sample descriptor are in meaning and sensory features, and the higher the degree of matching. The lower the semantic similarity score, the greater the difference between the two in terms of descriptive content and sensory performance, and the lower the degree of matching.

[0069] By calculating the degree of matching through semantic similarity models, the differences between descriptive words can be objectively and quantitatively judged from a semantic level, avoiding judgment bias caused by different human subjective understanding, and making difference recognition more standardized, automated and intelligent.

[0070] In the embodiments described in this specification, the quality assessment information only includes quality descriptive terms, and the quality standard sample information only includes the corresponding standard sample descriptive terms. The system compares the two types of descriptive terms one by one under each quality dimension to calculate their semantic or content matching degree. When the matching degree of the descriptive terms of any quality dimension (i.e., the first quality dimension) is less than a preset first threshold, the system determines that there is a significant quality difference in that dimension and marks that dimension as the target difference item. At the same time, the matching degree between the quality descriptive terms and the standard sample descriptive terms under that dimension is directly used as the difference information corresponding to the target difference item. This method can complete the difference identification without relying on quantitative scoring, which can perfectly adapt to the traditional tea evaluation scenario that only uses qualitative sensory descriptions and does not use digital scoring, further expanding the applicability of the solution. At the same time, by replacing human subjective judgment with objective rules of matching degree + threshold, it effectively avoids the ambiguity and subjectivity of judgment caused by differences in the understanding of descriptive terms, and ensures the standardization, objectivity and consistency of the difference identification process. In addition, its judgment logic is simple and intuitive, and the data processing is lightweight. It can quickly output stable and traceable difference results, providing a clear and reliable qualitative difference basis for subsequent quality assessment and supply chain optimization.

[0071] In another optional embodiment of this example, the quality assessment information includes a quality score, and the quality standard sample information includes a standard sample score. The quality scores across multiple quality dimensions and the standard sample scores are compared to determine the target differences between the tea to be evaluated and the standard tea across multiple quality dimensions, as well as the difference information corresponding to each target difference, including: The quality scores under multiple quality dimensions are compared with the standard sample scores to determine the score differences between the quality descriptors under multiple quality dimensions and the standard sample descriptors. If the difference between the quality score of the second quality dimension and the score of the standard sample is greater than the second threshold, then the second quality dimension is determined as the target difference item, wherein the second quality dimension is any one of the multiple quality dimensions. The difference between the quality score of the second quality dimension and the score of the standard sample is determined as the difference information corresponding to the target difference item.

[0072] The system calculates the score difference between the quality scores of multiple dimensions and the scores of the standard sample. The score difference is typically the absolute value of the difference between the two scores. For example, if the tea being evaluated scores 7 points in the "aroma intensity" dimension, and the standard sample tea scores 9 points, then the score difference = |7 - 9| = 2.

[0073] In addition, the system presets a maximum acceptable deviation range (second threshold). If the calculated difference exceeds this range, it indicates a non-negligible deviation in this dimension. For example, suppose the system sets the second threshold for the "aroma intensity" dimension to 1.5. In the example above, the difference 2 > 1.5, so the system will mark the "aroma intensity" dimension as a "target difference item".

[0074] The difference information is directly obtained using the calculated "rating difference" value. This is a very clear and explicit quantitative indicator.

[0075] Using the previous example, the system outputs the following difference information: Score Difference: 2 points. This precisely indicates that in the "Aroma Intensity" dimension, the tea to be evaluated is 2 points worse than the standard tea.

[0076] In this embodiment, by systematically comparing the quantitative scores of the tea to be evaluated and the standard sample across various quality dimensions, and automatically identifying significant differences based on preset thresholds, a fundamental shift in tea quality assessment from traditional manual judgment to digital and automated control is achieved. This solution transforms quality information into calculable numerical data. By accurately calculating the score difference and comparing it with a second threshold, it automatically filters out target difference items with significant deviations and generates quantitative difference information. This completely eliminates the volatility and inconsistency caused by subjective human factors during the evaluation process, improving the efficiency and reliability of quality assessment.

[0077] Step 106: Based on the difference dimension constraints corresponding to the tea to be evaluated, analyze at least one target difference item and the difference information corresponding to each target difference item to determine the quality evaluation result of the tea to be evaluated. The difference dimension constraints are configured based on the attributes of the tea to be evaluated and are used to indicate the constraint relationship between the difference item, the difference information and the quality evaluation result. The difference dimension constraints include single dimension constraints and / or total dimension constraints.

[0078] The difference dimension constraint is a rule system used to standardize the logical relationship between "difference items," "difference information," and "quality assessment results" in the tea quality assessment process. Its core is to clearly define, through pre-set specific and executable judgment criteria, what clear and standardized quality assessment results the assessment system should output when the tea to be assessed and the standard tea exhibit differences in a specific quality dimension (i.e., difference items, such as aroma, taste, and liquor color) that meet certain characteristics (i.e., difference information, including the number of difference dimensions, the specific numerical range of semantic similarity for each dimension, and the severity level of the difference). These results should be labeled as "qualified," "unqualified," "pending re-inspection," or a specific quality grade assessment and improvement suggestions. In short, this constraint essentially establishes a precise correspondence between "difference manifestations" and "assessment conclusions," ensuring the objectivity and consistency of the assessment results.

[0079] It should be understood that different types, grades, and origins of tea have different quality requirements and tolerance levels. For example, for a high-end pre-Qingming Longjing tea, the constraints will be very strict (e.g., no off-odors allowed, aroma intensity difference not exceeding 0.5 points). For a mass-produced pan-fired green tea, the constraints may be relatively lenient (e.g., a slight roasting flavor is allowed, aroma intensity difference can be tolerated up to 1.5 points). The system will automatically select which set of rules to use based on the attributes of the tea being evaluated.

[0080] In one optional implementation of this embodiment, the difference dimension constraints include single-item dimension constraints and / or total dimension constraints; based on the difference dimension constraints corresponding to the tea to be evaluated, the target difference items and the difference information corresponding to each target difference item are analyzed to determine the quality evaluation result of the tea to be evaluated, including: Based on single-dimensional constraints and / or total-dimensional constraints, analyze the target difference items and the difference information corresponding to each target difference item to determine the quality assessment result of the tea to be evaluated.

[0081] Among them, the single-dimensional constraint is a rigid judgment rule set separately for a certain independent quality dimension. It is used to control whether the tea has serious defects or fatal deviations in the core / key quality dimensions. If any dimension violates the constraint, the quality can be directly judged as substandard. It belongs to the "one-vote veto" rule to ensure that the tea will not have unacceptable problems in a single key dimension.

[0082] Example: Preset single-dimensional constraints are: If the degree of matching between the quality description of tea and the description of the standard sample in any one of the core dimensions of aroma or taste is less than 60%, or the absolute value of the score difference is greater than 5 points, then that dimension is judged to violate the single-dimensional constraint.

[0083] If the matching degree of the descriptive terms for the aroma dimension of the tea to be evaluated is only 45%, even if there are no significant differences in other dimensions, the quality evaluation result will be directly judged as unqualified because it violates the single-dimensional constraint conditions.

[0084] The overall dimension constraint is a comprehensive judgment rule set for the overall difference of all quality dimensions. It does not focus on the small deviation of a single dimension, but rather on the total number of target difference items, the overall score difference, the overall matching degree mean, and other overall indicators. It is used to avoid the overall quality decline caused by "small deviation in a single dimension, but the superposition of deviations in multiple dimensions". It belongs to the "overall control" category of rules.

[0085] Example: The preset total dimension constraint is: In all quality dimensions, the total number of target difference items is ≤3, and the sum of the absolute values ​​of the overall score differences across all dimensions is ≤8 points.

[0086] If the tea to be evaluated does not violate any single-item constraints in any dimension, but the four dimensions of appearance, liquor color, leaf residue, and taste are all target difference items, and the total number of target difference items (4) exceeds the upper limit of the total dimension constraints (3), then the total dimension constraints are violated, and the quality evaluation result is judged as unsuccessful.

[0087] In the embodiments described in this specification, each quality dimension is independently and rigorously reviewed through single-dimensional constraints. This accurately identifies fatal flaws such as off-flavors or missing characteristic aromas, ensuring that the tea is free of substantial defects in any single key dimension, thus maintaining the bottom line of quality control. Building upon this, the overall dimensional constraints take a macro-level approach, comprehensively evaluating the overall performance of the tea. By setting rules such as total score tolerance and upper limits on the total number of discrepancies, it effectively avoids overall quality degradation caused by the accumulation of minor deviations across multiple dimensions, ensuring the stability and harmony of the product's flavor.

[0088] In one optional implementation of this embodiment, the quality assessment information includes quality descriptive terms, and the difference dimension constraints include single-item dimension constraints. Based on the single-item dimension constraints, the target difference items and the difference information corresponding to each target difference item are analyzed to determine the quality assessment result of the tea to be evaluated, including: If any quality descriptor for any target difference item contains a descriptor that matches the first keyword, then the quality assessment result for the tea to be evaluated is determined to be unsatisfactory; or, If the degree of match between the quality descriptor of any target difference item and the standard sample descriptor is less than the third threshold, then the quality assessment result of the tea to be evaluated is determined to be unsuccessful; or, If the target difference items include key quality dimensions, then the quality assessment result of the tea to be evaluated is determined to be unsatisfactory.

[0089] Among them, the keyword-triggered veto mechanism (absolute defect identification) is the core implementation method of the veto logic. Its specific rule is: if a descriptive word matching the system's predefined set of "first keywords" is detected in the quality descriptive words corresponding to any target difference item, the quality evaluation result of the tea to be evaluated is directly determined to be unsuccessful.

[0090] This mechanism relies on a predefined "negative terminology library" (i.e., a set of "primary keywords") within the system. The terms in this library represent unacceptable, fatal flaws in the tea's quality. The system scans all target discrepancies (i.e., quality dimensions that deviate from the standard tea sample) for the quality descriptions assigned to the tea by the reviewers. If a description is found to match a keyword in the negative terminology library in any dimension, the system immediately triggers a "fail" decision.

[0091] This negative keyword database can be configured in two ways: Global thesaurus: A unified thesaurus applicable to all quality dimensions. The appearance of such words in any dimension will trigger a rejection.

[0092] Dimension-specific terminology: Set up unique negative terminology for different quality dimensions (such as aroma, taste, appearance, etc.) to achieve more refined control.

[0093] Example explanation: Example of a global keyword database: The first keyword set can include musty smell, sour smell, smoky smell, no aroma, off-odor, etc. Regardless of which quality dimension these words appear in (e.g., "musty smell" in "taste" or "sour smell" in "aroma"), the system will immediately determine "not passed".

[0094] Example of a dimension-specific thesaurus: For the dimensions "aroma" and "taste," the specific primary keywords can include no characteristic aroma, no characteristic taste, off-odor, and off-taste. The system will focus on scanning descriptive words for these two dimensions, and will immediately reject any matching terms.

[0095] Among them, the low similarity triggers rejection (serious inconsistency), and its specific rule is: if the degree of matching between the quality descriptor of any target difference item and the standard sample descriptor is less than the third threshold, then the quality evaluation result of the tea to be evaluated is determined to be unsuccessful.

[0096] This rule focuses on the degree of deviation of the tea being evaluated from the standard tea sample. Even if the tea being evaluated has no absolute defects, it is unacceptable if its flavor characteristics differ too much from those of the standard tea sample.

[0097] The system calculates the semantic similarity (matching degree) between the descriptive terms of the tea to be evaluated and the descriptive terms of the standard sample on the problematic dimension (difference item); this similarity is then compared with a preset third threshold (e.g., similarity <50%). If it is below this threshold, it indicates a serious inconsistency in flavor, triggering a rejection.

[0098] For example, the standard tea sample is described as having a rich aroma, while the tea to be evaluated is described as having a mild aroma. The system calculates that the similarity between the two is extremely low (e.g., 30%), and since 30% < 50% (the third threshold), it is judged as "failed".

[0099] This rule ensures the stability and consistency of product flavor, guaranteeing that consumers experience the same taste every time they drink the product.

[0100] Among them, the key dimension difference triggers rejection (core trait missing). The specific rule is: if the target difference item includes a key quality dimension, the quality assessment result of the tea to be evaluated is determined to be unsuccessful.

[0101] Specifically, the system checks the list of target discrepancies; it then checks whether the list contains items that are predefined as "key quality dimensions," and if so, it immediately triggers a rejection.

[0102] Example: For jasmine tea, the intensity and purity of the jasmine fragrance are the soul, therefore the "aroma" dimension is defined as a key quality dimension. As long as the "aroma" dimension appears in the target difference item list (i.e., there is a difference from the standard tea sample), regardless of the size of the difference, it will be directly judged as "fail".

[0103] In the embodiments described in this specification, efficient and accurate automated decision-making for tea quality assessment is achieved by establishing single-dimensional constraints based on quality descriptive terms. First, a first keyword matching mechanism (such as "moldy smell" or "sour smell") quickly identifies absolute defects, ensuring that teas with serious quality problems are directly rejected. Second, semantic similarity calculation and comparison with a third threshold effectively screen out products whose flavor characteristics significantly deviate from the standard sample, ensuring the stability and consistency of the flavor profile. Furthermore, by setting key quality dimensions (such as the "floral aroma" characteristic of jasmine tea), the core value attributes of the product are ensured to remain intact; any deviation in this dimension triggers rejection, thus precisely safeguarding the product's market competitiveness. This rule system can be executed without complex calculations, significantly improving quality control efficiency, eliminating the volatility of subjective human judgment, and providing objective, reliable, and traceable automated decision support for tea quality control.

[0104] In another optional implementation of this embodiment, the quality assessment information includes a quality score, and the difference dimension constraints include single-item dimension constraints. Based on the single-item dimension constraints, the target difference items and the difference information corresponding to each target difference item are analyzed to determine the quality assessment result of the tea to be evaluated, including: If the quality score of any target difference item matches the standard score less than the fourth threshold, the quality assessment result of the tea to be evaluated is determined to be unsuccessful.

[0105] The fourth threshold defines the maximum acceptable difference in scores for the system, used to determine whether a single dimension of quality is up to standard. If the difference between the score of the tea to be evaluated and the score of the standard tea in a certain dimension reaches or exceeds this threshold, the rejection mechanism is directly triggered, and the quality evaluation result is deemed unsatisfactory.

[0106] For example, the standard tea sample scored 8 points in the "aroma" dimension: If the tea to be evaluated scores 7 or 9 points in this dimension (with an absolute difference of 1 point), and does not reach the fourth threshold (e.g., set to 2 points), it is considered an acceptable deviation and will not trigger a rejection. If the tea to be evaluated scores 6 or 10 points (with an absolute difference of 2 points), reaching or exceeding the fourth threshold, the system will immediately determine that the tea quality fails.

[0107] In one optional implementation of this embodiment, the difference dimension constraint includes a total dimension constraint, whereby the number of difference items is greater than a threshold. Based on the total dimension constraint, the target difference items and / or the difference information corresponding to each target difference item are analyzed to determine the quality assessment result of the tea to be evaluated, including: If the number of detected differences in the target difference item exceeds the quantity threshold, the quality assessment result of the tea to be evaluated is determined.

[0108] In this embodiment of the specification, a quantity threshold is first preset for the evaluation process (i.e., the maximum number of quality dimensions that are allowed to differ between the tea to be evaluated and the standard tea), which is used as the overall dimension constraint. Then, the total number of target difference items that are significantly different between the tea to be evaluated and the standard tea is counted, and this number is compared with the preset quantity threshold. Once the number of target difference items exceeds the quantity threshold, the quality evaluation result of the tea to be evaluated is directly determined according to this rule. In this way, standardized and quantitative quality control is achieved from the overall dimension difference quantity, ensuring that the evaluation results are objective and consistent.

[0109] Assuming that the system's preset total dimension constraint condition has a quantity threshold of 3 (this threshold can be flexibly configured based on the attributes of the tea to be evaluated, such as tea type, grade, etc.); after the system performs difference analysis, it determines the target difference item list as "aroma intensity", "flavor thickness", "liquor color brightness" and "leaf uniformity", and after statistics, the number of difference items is 4; at this time, a logical judgment is made: the number of difference items (4) > the quantity threshold (3), the total dimension constraint condition is met, that is, the tea to be evaluated violates the total dimension constraint condition; once the total dimension constraint condition is met, the system will directly determine the quality evaluation result of the tea to be evaluated as "not passed".

[0110] Conversely, if the number of target differences counted by the system is 2 (for example, only the two dimensions of "aroma intensity" and "flavor thickness" have significant differences), the logical judgment is: the number of differences (2) < the number threshold (3), and the total dimension constraint is not valid; when the constraint is not valid, the "veto rule" will not be triggered, and the system will continue to combine other constraints (such as single dimension constraints) to comprehensively determine the quality of the tea to be evaluated.

[0111] In the embodiments described in this specification, the introduction of a total dimension constraint effectively achieves macro-level control over systematic deviations in tea quality. It can accurately identify the risk of overall quality degradation caused by the accumulation of minor defects across multiple dimensions: when the number of target discrepancies exceeds a preset threshold, the system automatically determines failure, thus preventing the problem of "single dimension meeting standards but overall quality instability" that may occur in traditional assessments. This mechanism simulates the comprehensive judgment thinking in expert review, not only improving the comprehensiveness and reliability of quality assessment but also significantly increasing decision-making efficiency, especially suitable for rapid quality control in large-scale production scenarios. Simultaneously, the flexible configurability of the quantity threshold allows the system to adapt to the differentiated quality requirements of different categories and grades of tea, ensuring product flavor consistency and stability while providing clear data guidance and decision-making basis for production process optimization.

[0112] In one optional implementation of this embodiment, after obtaining the quality assessment information of the tea to be evaluated in multiple quality dimensions and the quality standard information of the standard tea in multiple quality dimensions, the method further includes: Based on the quality assessment information of the tea to be evaluated in multiple quality dimensions, the comprehensive assessment information of the tea to be evaluated in multiple quality dimensions is determined. Obtain comprehensive standard sample information; Based on comprehensive evaluation information and comprehensive standard sample information, the comprehensive differences between the tea to be evaluated and the standard tea are determined; Accordingly, based on the constraint conditions of the difference dimensions corresponding to the tea to be evaluated, the target difference items and the difference information corresponding to each target difference item are analyzed to determine the quality evaluation results of the tea to be evaluated, which also includes: Based on the difference dimension constraints and comprehensive difference constraints corresponding to the tea to be evaluated, the comprehensive difference information, target difference items, and difference information corresponding to each target difference item are analyzed to determine the quality evaluation result of the tea to be evaluated. Among them, the comprehensive difference constraints are configured based on the attributes of the tea to be evaluated and are used to indicate the constraint relationship between the comprehensive difference information and the quality evaluation result.

[0113] The system calculates a single comprehensive score or grade representing the overall quality of the tea based on its quality assessment information (such as quality scores or quality descriptive terms) across all individual quality dimensions using a pre-defined algorithm model (such as weighted average, multinomial calculation, machine learning model, etc.). This is the "comprehensive assessment information".

[0114] In this scheme, comprehensive standard sample information specifically refers to the digital and structured representation of the overall quality level of a standard tea sample (standard sample). Comprehensive standard sample information can be condensed into a calculable and comparable macro-indicator (usually a score or grade) by using a pre-defined, unified algorithm model (such as a weighted comprehensive score or a multi-index fusion model) to encapsulate all scores or descriptive terms of the standard sample across various main and sub-dimensions. For example, the comprehensive standard sample information of a premium Longjing tea might be quantified as "Comprehensive Quality Score: 92 points" or "Grade: A".

[0115] In practice, the comprehensive standard sample information can either allow users to directly upload a predetermined comprehensive score for the standard tea sample, or the system can use relevant algorithms to calculate and generate the score in real time from the underlying dimensional data of the standard tea sample.

[0116] By comparing the two, the overall difference is calculated, which is the "comprehensive difference information". This is usually a numerical value, such as the total score difference (for example: if the tea to be evaluated has a comprehensive score of 85 points and the standard sample has a comprehensive score of 90 points, then the comprehensive difference information is -5 points).

[0117] The comprehensive difference constraint refers to an automated judgment rule set based on the overall quality difference between the tea to be evaluated and the standard tea. It can take the form of: the comprehensive difference information exceeding a preset overall deviation threshold. For example, a failure is triggered if the absolute value of the comprehensive difference is greater than 5 points. It can also take the form of: the comprehensive evaluation grade being lower than the standard sample by a certain number of grades. For example, a grade difference exceeding two grades results in a failure. Furthermore, it can take the form of: the comprehensive difference information exceeding the acceptable distribution range defined based on historical data or quality standards (e.g., exceeding the mean ± 2 standard deviations).

[0118] In the final decision-making process, the system's judgment logic is upgraded to a multi-dimensional adjudication mechanism. It no longer considers micro- or macro-level differences in isolation, but rather requires the simultaneous satisfaction of two sets of rules: Existing variance dimension constraints: Continue to check for unacceptable single-dimensional defects (veto) or an excessive number of variance items.

[0119] The newly added constraint on overall differences checks whether the overall differences are within an acceptable range. For example, the rule might stipulate: "If the overall difference is greater than 5 points, it will be directly judged as a failure."

[0120] In the embodiments described in this specification, the introduction of a comprehensive evaluation and multi-dimensional adjudication mechanism significantly improves the systematicness, accuracy, and intelligent decision-making level of tea quality evaluation. It achieves comprehensive quality control from local to global perspectives: the system uses algorithms to fuse data from various dimensions to generate comprehensive evaluation information and compares it with a comprehensive standard sample, thereby effectively identifying overall quality deviations in the tea to be evaluated and preventing systemic risks caused by the superposition of minor deviations in multiple dimensions; simultaneously, through the synergistic application of comprehensive difference constraints and original difference dimension constraints, a dual guarantee mechanism is constructed that combines micro-defect detection and macro-consistency judgment. This mechanism can accurately capture fatal defects in a single dimension while strictly controlling quality consistency based on overall differences, greatly improving the comprehensiveness and reliability of the judgment results.

[0121] In the above embodiments, multiple quality dimensions are primary dimensions, meaning that there are no sub-dimensions under each quality dimension. In another optional embodiment, the quality dimension system can adopt a hierarchical structure, that is, it includes multiple main dimensions (such as aroma, taste, appearance, etc.), and each main dimension can be further subdivided into several sub-dimensions (for example, the main dimension of "aroma" can include sub-dimensions such as "characteristic aroma", "purity", and "intensity").

[0122] Under this hierarchical dimensional system: the target difference item specifically refers to the sub-dimension that deviates significantly from the standard tea sample under each main dimension.

[0123] The difference dimension constraints also include individual dimension constraints and total dimension constraints, but their specific implementation needs to adapt to the dimension hierarchy structure: Single-dimensional constraint: Applicable to a single sub-dimension. This constraint determines whether a difference in a specific sub-dimension constitutes a veto. For example, a veto rule can be set for the "purity" sub-dimension; if "exotic gases" are detected, it will be directly deemed a failure.

[0124] Overall Dimension Constraint: Applicable to the main dimension or the entire evaluation system. This constraint limits the number of sub-dimensions that can exhibit discrepancies from a global perspective. The judgment is based on the number of discrepancies, i.e., counting the number of discrepancies among all sub-dimensions under a given main dimension, or comparing the total number of discrepancies across all main dimensions to a preset threshold. If the number of discrepancies exceeds the threshold, the evaluation fails.

[0125] See Figure 2 , Figure 2 A schematic diagram of a tea end-to-end quality control and supply chain optimization system provided according to an embodiment of this specification is shown.

[0126] like Figure 2 As shown, the system includes a supply chain management terminal 202 and a quality testing terminal 204, constructing a closed-loop system of "supply chain sampling → quantitative evaluation at the testing terminal → result feedback → precise improvement of the supply chain → secondary sampling verification". It can solve the problems of disconnect between the tea supply chain and quality testing, inability to accurately locate quality problems, and lack of basis for improvement, and realize "driving supply chain optimization with test data and ensuring quality stability with supply chain optimization".

[0127] The specific workflow of this system is as follows: Based on actual control needs such as raw material warehousing inspection, production batch sampling inspection, and finished product quality verification, the supply chain management end determines the tea samples to be evaluated and formally initiates a quality testing request. At the same time, it enters the basic traceability information of the batch of samples, including but not limited to: raw material origin, raw material supplier, harvesting time, processing parameters, processing time, storage environment information, batch number, and personnel submitting the samples for testing. This completes the initialization of the basic traceability data. The tea samples to be evaluated and their corresponding basic traceability information are then directly submitted to the quality testing end, which simultaneously records the sample submission time, personnel submitting the samples for testing, and other traceability information, thus achieving traceability of the sample submission process.

[0128] The quality testing terminal receives the tea samples to be evaluated and the corresponding traceability information, and conducts a standardized quality evaluation on the tea samples to be evaluated. Specifically, it obtains the quality evaluation information (including quality description words and / or quality scores) of the tea to be evaluated in multiple quality dimensions, and the quality standard sample information (including standard sample description words and / or standard sample scores) of the standard sample tea in multiple quality dimensions; based on the quality evaluation information and the corresponding quality standard sample information in multiple quality dimensions, through methods such as dimension-by-dimension comparison, semantic similarity calculation (for description words), and difference threshold determination (for scores), it determines at least one target difference item between the tea to be evaluated and the standard sample tea in multiple quality dimensions and the difference information corresponding to each target difference item (such as the matching degree of description words, score difference, etc.); based on the difference dimension constraint conditions corresponding to the tea to be evaluated (including single dimension constraint conditions and / or total dimension constraint conditions), it analyzes at least one target difference item and the difference information corresponding to each target difference item to determine the quality evaluation result of the tea to be evaluated (such as pass / fail / need re-evaluation, quality score), where the difference dimension constraint conditions are configured based on the attributes (category, grade, etc.) of the tea to be evaluated and are used to indicate the constraint relationship between the difference items, difference information and the quality evaluation result. At the same time, the quality testing terminal binds all the data of this evaluation (evaluation time, testing personnel, scores / description words of each dimension, target difference items, difference information, quality evaluation result, improvement suggestions, etc.) with the traceability information of the sample, ensuring that each quality data can correspond to a specific sample and specific batch, and realizing the accurate traceability of quality problems.

[0129] Based on the quality evaluation result, target difference items, difference information and the complete traceability file, the quality testing terminal determines the quality feedback information (which can also include targeted improvement suggestions), and directly feeds back the quality feedback information to the supply chain management terminal, ensuring that the supply chain management terminal can not only quickly obtain quality problems, but also accurately locate the root cause of the problems through the traceability file.

[0130] The supply chain management terminal receives the quality feedback information sent by the quality testing terminal, traces the specific links of the quality problem (such as raw material problems, process problems, warehousing problems) in combination with the traceability file, and at the same time takes into account the actual situation of the supply chain (cost, production capacity, supplier resources), and implements targeted improvements in the corresponding links. For example: Raw material procurement link: If the test feedback is "insufficient aroma of raw materials", confirm the raw material origin or supplier problem in combination with the traceability information, optimize the raw material procurement standard, replace low-quality suppliers, and enter the improvement information such as procurement standard adjustment and supplier replacement into the traceability file.

[0131] Production and processing link: If the test feedback is "excessive bitterness and astringency of taste", adjust the process parameters such as fixation, fermentation, scenting, etc. in combination with the processing process parameters in the traceability information, record the details of process adjustment and update the traceability file.

[0132] In the warehousing and logistics process: if the test results indicate "fragrance loss after warehousing", the warehousing environment data in the traceability information will be combined to adjust the warehousing temperature and humidity, shorten the turnover cycle, and update the traceability records of warehousing management in a timely manner.

[0133] After the improvements are completed, the supply chain management end initiates a second sample submission request, directly submitting the improved samples and traceability updates of the improved processes to the quality testing end to verify the effectiveness of the improvements.

[0134] The quality testing end receives the improved second sample from the supply chain management end, repeats the above standardized evaluation process, focuses on comparing the differences between the current evaluation results and the first evaluation results, and outputs an evaluation of the improvement effect (such as "the aroma score has increased by 3 points compared to the last time, the improvement is effective" and "the bitterness of the taste has been reduced to the standard sample range, the improvement meets the standard"). At the same time, the second evaluation data and the improvement effect conclusion are linked to the traceability information, the complete traceability file is updated, and the improvement effect evaluation and the updated traceability file are fed back to the supply chain management end.

[0135] After receiving feedback on the improvement effects of the secondary evaluation and the complete traceability file, the supply chain management end proceeds with follow-up work in two ways: If the improvement is effective, the optimized procurement standards, process parameters, and warehousing conditions are solidified as supply standards and simultaneously entered into the traceability file as the benchmark for subsequent batches of tea control, achieving a standardized upgrade of the supply chain; if the improvement does not meet the standards, the root cause of the problem is analyzed again based on the traceability file, the improvement plan is adjusted, and the "improvement-second sampling-evaluation" process is repeated until the quality meets the standards. At the same time, the supply chain management end summarizes the traceability data, evaluation data, and improvement records of all batches to form a quality control ledger for subsequent review and summary, quality traceability, and new product development, further improving the stability of tea quality and the level of refined control, and achieving a closed-loop iteration of full-chain quality control and supply chain optimization.

[0136] The following is in conjunction with the appendix Figure 3 Taking the application of the tea evaluation method provided in this manual in a tea quality testing scenario as an example, the tea evaluation method will be further explained. Among other things, Figure 3 The present specification shows a flowchart of a tea evaluation method according to an embodiment, which includes the following steps.

[0137] Step 302: Obtain the quality assessment information of the tea to be evaluated on multiple sub-dimensions under each main dimension, and the quality standard information of the standard tea on multiple sub-dimensions under each main dimension. The quality assessment information includes quality scores and / or quality descriptive terms, and the quality standard information includes standard scores and / or standard descriptive terms. The main dimensions and multiple sub-dimensions under each main dimension are configured based on the attributes of the tea to be evaluated.

[0138] For example, the quality dimension includes five main dimensions: appearance, liquor color, aroma, taste, and infused leaf residue. Each main dimension can be further subdivided into several quantifiable evaluation sub-dimensions. Examples of sub-dimensions for each main dimension are as follows.

[0139] Appearance: further subdivided into shape, strands, tenderness, color, and uniformity; Soup color: further subdivided into color and brightness; Aroma: Subdivided into characteristic aroma (including aroma type, purity, intensity, and persistence) and unusual odor (including type and intensity of unusual odor). Taste: Subdivided into characteristic taste (including flavor profile, aroma, and consistency), body (including concentration and thickness), bitterness (including bitterness, astringency, and the speed at which astringency dissipates), and unusual taste (including off-flavor type and intensity). Leaf base: further subdivided into tenderness, color (including color and brightness), and uniformity.

[0140] The process of obtaining the quality assessment information of the tea to be evaluated in multiple sub-dimensions under each main dimension, and the quality standard information of the standard tea in multiple sub-dimensions under each main dimension, can refer to the relevant description of step 102 in the above embodiment.

[0141] Step 304: Based on the quality assessment information of multiple sub-dimensions under each main dimension and the corresponding quality standard sample information, determine the target difference items between the tea to be evaluated and the standard tea, as well as the difference information corresponding to each target difference item.

[0142] The specific implementation of step 304 can refer to step 104 of the above embodiment.

[0143] For example, if the difference between the quality score of any sub-dimension and the standard score exceeds a set threshold, then the sub-dimension is determined to be the target difference item under the corresponding main dimension, and the score difference is the difference information of the target difference item under the corresponding main dimension.

[0144] Alternatively, if the semantic similarity between the quality descriptor of any sub-dimension and the standard descriptor is lower than a set threshold, then that sub-dimension is determined to be the target difference item under the corresponding main dimension, and the score difference is the difference information of the target difference item under the corresponding main dimension.

[0145] Step 306: Based on the quality assessment information of the tea to be evaluated in multiple sub-dimensions under each main dimension, determine the comprehensive assessment information of the tea to be evaluated and obtain the comprehensive standard sample information; based on the comprehensive assessment information and the comprehensive standard sample information, determine the comprehensive difference information between the tea to be evaluated and the standard sample tea.

[0146] For example, based on the quality scores of the tea to be evaluated on multiple sub-dimensions under each main dimension, a weighted sum is performed in combination with the corresponding sub-dimension weights to obtain the main dimension score for each main dimension. The comprehensive evaluation information system then calculates the comprehensive evaluation information based on the main dimension scores and their weights, and finally performs a layer-by-layer weighted summation of all main dimension scores with pre-set main dimension weights to obtain a comprehensive quality score that fully reflects the overall quality level. This comprehensive quality score is then used as the comprehensive evaluation information for the tea to be evaluated.

[0147] At the same time, the comprehensive standard sample score corresponding to the standard tea sample is obtained (the calculation method is the same as the above method, using the same sub-dimension weights and main dimension weights), and the difference between the comprehensive evaluation information and the comprehensive standard sample information is determined as the comprehensive difference information between the tea to be evaluated and the standard tea sample.

[0148] Step 308: Based on the difference dimension constraints and comprehensive difference constraints corresponding to the tea to be evaluated, analyze the comprehensive difference information, target difference items, and the difference information corresponding to each target difference item to determine the quality evaluation result of the tea to be evaluated.

[0149] The specific implementation of step 308 can refer to step 106 of the above embodiment.

[0150] For example, the single-dimensional constraint is as follows: Scoring Difference: If the score difference between the tea to be evaluated and the standard tea in a certain main dimension exceeds the set threshold (such as ±3 points or ±5 points), then that main dimension is considered a difference item.

[0151] Aroma and taste: The maximum allowable difference is ±5 points; anything exceeding this will be considered a failure.

[0152] Appearance, soup color, and leaf residue: The maximum allowable difference is ±3 points; anything exceeding this will be considered a failure.

[0153] Descriptor Difference: If the semantic similarity between the quality descriptor and the standard descriptor is lower than the threshold, it is considered as failing.

[0154] Keyword trigger: If a set negative keyword appears in the description, it will be considered as not passing.

[0155] The comprehensive dimensional constraints are as follows: (1) Overall score difference: Calculate the overall score difference between the tea to be evaluated and the standard tea; if the overall score difference is... If the score exceeds the set threshold (e.g., ±5 points), it is considered a failure.

[0156] (2) Number of Difference Dimensions: If the number of sub-dimensions that deviate significantly from the standard sample under a certain main dimension exceeds the preset number threshold corresponding to that main dimension, the quality assessment result is deemed unsuccessful. For example, the preset number threshold corresponding to the main dimensions of aroma and taste is 4, and the preset number threshold corresponding to the main dimensions of appearance, soup color, and leaf residue is 2. If the number of deviation sub-dimensions under any main dimension exceeds the corresponding threshold, it is deemed unsuccessful.

[0157] Comprehensive judgment logic: Grade A: Overall score difference ≤ ±3 points, and the number of sub-dimensions of difference in key dimensions (aroma, taste) ≤ 4, and the number of sub-dimensions of difference in other dimensions (appearance, soup color, infused leaves) ≤ 2.

[0158] Grade A: Overall score difference ≤ ±5 points. If there are any doubts, the assessment should be based on the application test results.

[0159] Grade B: If the overall score difference is greater than ±5 points, it is directly judged as failing.

[0160] Results determination: If the tea to be evaluated meets the comprehensive difference constraint conditions and the single-dimensional difference is within the allowable range, it is judged as passing (Grade A or Grade A-).

[0161] If any condition is not met, the result is deemed a failure (Grade B).

[0162] By following the steps above, the quality of the tea to be evaluated can be comprehensively assessed, ensuring the accuracy and consistency of the evaluation results.

[0163] Corresponding to the above method embodiments, this specification also provides embodiments of a tea evaluation device. Figure 4 A schematic diagram of a tea evaluation device according to one embodiment of this specification is shown. Figure 4 As shown, the device includes: The acquisition module 402 is configured to acquire quality assessment information of the tea to be evaluated in multiple quality dimensions, and quality standard information of the standard tea in multiple quality dimensions, wherein the multiple quality dimensions are configured based on the attributes of the tea to be evaluated. The first determining module 404 is configured to determine at least one target difference item between the tea to be evaluated and the standard tea in multiple quality dimensions and the difference information corresponding to each target difference item based on quality assessment information of multiple quality dimensions and corresponding quality standard sample information. The second determining module 406 is configured to analyze at least one target difference item and the difference information corresponding to each target difference item based on the difference dimension constraint conditions corresponding to the tea to be evaluated, and determine the quality evaluation result of the tea to be evaluated. The difference dimension constraint conditions are configured based on the attributes of the tea to be evaluated and are used to indicate the constraint relationship between the difference item, the difference information and the quality evaluation result. The difference dimension constraint conditions include single dimension constraint conditions and / or total dimension constraint conditions.

[0164] Optionally, the quality assessment information includes quality descriptive terms and / or quality scores, and the quality standard sample information includes standard sample descriptive terms and / or standard sample scores; the first determining module 404 is further configured to: compare the quality descriptive terms and / or quality scores under multiple quality dimensions, as well as the standard sample descriptive terms and / or standard sample scores, to determine the target difference items between the tea to be evaluated and the standard tea in multiple quality dimensions and the difference information corresponding to each target difference item.

[0165] Optionally, the quality assessment information includes quality descriptive terms, and the quality standard sample information includes standard sample descriptive terms; the first determining module 404 is further configured to: The quality descriptors under multiple quality dimensions are compared with the standard descriptors to determine the degree of matching between the quality descriptors under multiple quality dimensions and the standard descriptors. If the degree of matching between the quality descriptors of the first quality dimension and the standard descriptors is less than a first threshold, then the first quality dimension is determined as the target difference item, where the first quality dimension is any one of the multiple quality dimensions. The degree of matching between the quality descriptors of the first quality dimension and the standard descriptors is determined as the difference information corresponding to the target difference item.

[0166] Optionally, the quality assessment information includes a quality score, and the quality standard sample information includes standard sample descriptive terms; the first determining module 404 is further configured to: The quality scores under multiple quality dimensions are compared with the standard scores to determine the score difference between the quality descriptors under multiple quality dimensions and the standard descriptors. If the score difference between the quality score of the second quality dimension and the standard score is greater than a second threshold, the second quality dimension is determined as the target difference item. The second quality dimension can be any one of the multiple quality dimensions. The score difference between the quality score of the second quality dimension and the standard score is determined as the difference information corresponding to the target difference item.

[0167] Optionally, the quality assessment information includes quality descriptive terms, and the difference dimension constraints include single-item dimension constraints; the second determination module 406 is further configured as follows: If any quality descriptor of a target difference item contains a descriptor that matches the first keyword, the quality assessment result of the tea to be evaluated is determined to be unsuccessful; or, if the degree of matching between the quality descriptor of any target difference item and the standard sample descriptor is less than the third threshold, the quality assessment result of the tea to be evaluated is determined to be unsuccessful; or, if the target difference item includes a key quality dimension, the quality assessment result of the tea to be evaluated is determined to be unsuccessful.

[0168] Optionally, the difference dimension constraint includes a total dimension constraint, whereby the number of difference items is greater than a threshold; the second determining module 406 is further configured to: If the number of detected differences in the target difference item exceeds the quantity threshold, the quality assessment result of the tea to be evaluated is determined.

[0169] Optionally, the tea evaluation device further includes a third determining module, configured to, after acquiring quality evaluation information of the tea to be evaluated in multiple quality dimensions and quality standard information of the standard tea in multiple quality dimensions, determine comprehensive evaluation information of the tea to be evaluated in multiple quality dimensions based on the quality evaluation information of the tea to be evaluated in multiple quality dimensions; acquire comprehensive standard information; and determine comprehensive difference information between the tea to be evaluated and the standard tea based on the comprehensive evaluation information and the comprehensive standard information. Accordingly, the second determining module 406 is further configured as follows: Based on the difference dimension constraints and comprehensive difference constraints corresponding to the tea to be evaluated, the comprehensive difference information, target difference items, and difference information corresponding to each target difference item are analyzed to determine the quality evaluation results of the tea to be evaluated.

[0170] Optionally, the acquisition module 402 is further configured as follows: The tea evaluation interface displays evaluation controls for multiple quality dimensions. These controls include preset quality information and / or input controls. The preset quality information is configured based on the attributes of the tea to be evaluated. In response to the evaluation operation of the evaluation control under the target quality dimension, the quality evaluation information under the target quality dimension is determined based on the preset quality information selected by the evaluation operation and / or the input quality information, wherein the target quality dimension is any one of multiple quality dimensions; In response to the completion event in the tea evaluation interface, obtain quality evaluation information for each quality dimension.

[0171] The above is a schematic diagram of a tea evaluation device according to this embodiment. It should be noted that the technical solution of this tea evaluation device and the technical solution of the tea evaluation method described above belong to the same concept. For details not described in detail in the technical solution of the tea evaluation device, please refer to the description of the technical solution of the tea evaluation method described above.

[0172] Figure 5 A structural block diagram of a computing device 500 according to one embodiment of this specification is shown. The components of the computing device 500 include, but are not limited to, a memory 510 and a processor 520. The processor 520 is connected to the memory 510 via a bus 530, and a database 550 is used to store data.

[0173] The computing device 500 also includes an access device 540, which enables the computing device 500 to communicate via one or more networks 560. Examples of these networks include Public Switched Telephone Network (PSTN), Local Area Network (LAN), Wide Area Network (WAN), Personal Area Network (PAN), or combinations of communication networks such as the Internet. The access device 540 may include one or more of any type of wired or wireless network interface (e.g., a network interface card (NIC)), such as an IEEE 802.11 Wireless Local Area Network (WLAN) wireless interface, a Wi-MAX (Worldwide Interoperability for Microwave Access) interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, or a Near Field Communication (NFC) interface.

[0174] In one embodiment of this specification, the above-described components of the computing device 500 and Figure 5 Other components, not shown, can also be connected to each other, for example, via a bus. It should be understood that... Figure 5 The block diagram of the computing device shown is for illustrative purposes only and is not intended to limit the scope of this specification. Those skilled in the art can add or replace other components as needed.

[0175] The computing device 500 can be any type of stationary or mobile computing device, including mobile computers or mobile computing devices (e.g., tablet computers, personal digital assistants, laptop computers, notebook computers, netbooks, etc.), mobile phones (e.g., smartphones), wearable computing devices (e.g., smartwatches, smart glasses, etc.) or other types of mobile devices, or stationary computing devices such as desktop computers or personal computers (PCs). The computing device 500 can also be a mobile or stationary server.

[0176] The processor 520 is used to execute the following computer-executable instructions, which, when executed by the processor, implement the steps of the above-described tea evaluation method.

[0177] The above is a schematic diagram of a computing device according to this embodiment. It should be noted that the technical solution of this computing device and the technical solution of the tea evaluation method described above belong to the same concept. For details not described in detail in the technical solution of the computing device, please refer to the description of the technical solution of the tea evaluation method described above.

[0178] An embodiment of this specification also provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the tea evaluation method described above.

[0179] The above is an illustrative scheme of a computer-readable storage medium according to this embodiment. It should be noted that the technical solution of this storage medium and the technical solution of the tea evaluation method described above belong to the same concept. For details not described in detail in the technical solution of the storage medium, please refer to the description of the technical solution of the tea evaluation method described above.

[0180] An embodiment of this specification also provides a computer program, wherein when the computer program is executed in a computer, it causes the computer to perform the steps of the above-described tea evaluation method.

[0181] The above is an illustrative example of a computer program according to this embodiment. It should be noted that the technical solution of this computer program and the technical solution of the tea evaluation method described above belong to the same concept. Details not described in detail in the technical solution of the computer program can be found in the description of the technical solution of the tea evaluation method described above.

[0182] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0183] The computer instructions include computer program code, which may be in the form of source code, object code, executable file, or certain intermediate forms. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium may be appropriately added or removed according to the requirements of patent practice. For example, in some regions, according to patent practice, computer-readable media may not include electrical carrier signals and telecommunication signals.

[0184] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments in this specification are not limited to the described order of actions, because according to the embodiments in this specification, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the embodiments in this specification.

[0185] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0186] The preferred embodiments disclosed above are merely illustrative of this specification. The optional embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the embodiments described herein. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the embodiments, thereby enabling those skilled in the art to better understand and utilize this specification. This specification is limited only by the claims and their full scope and equivalents.

Claims

1. A method for evaluating tea quality, characterized in that, include: The quality assessment information of the tea to be evaluated in multiple quality dimensions, and the quality standard information of the standard tea in the multiple quality dimensions, wherein the multiple quality dimensions are configured based on the attributes of the tea to be evaluated; Based on the quality assessment information of the multiple quality dimensions and the corresponding quality standard sample information, at least one target difference item between the tea to be evaluated and the standard tea in the multiple quality dimensions and the difference information corresponding to each target difference item are determined. Based on the difference dimension constraints corresponding to the tea to be evaluated, the at least one target difference item and the difference information corresponding to each target difference item are analyzed to determine the quality evaluation result of the tea to be evaluated. The difference dimension constraints are configured based on the attributes of the tea to be evaluated and are used to indicate the constraint relationship between the difference item, the difference information and the quality evaluation result. The difference dimension constraints include single dimension constraints and / or total dimension constraints.

2. The method according to claim 1, characterized in that, The quality assessment information includes quality descriptive terms and / or quality scores, and the quality standard sample information includes standard sample descriptive terms and / or standard sample scores; Based on the quality assessment information of the multiple quality dimensions and the corresponding quality standard sample information, at least one target difference item between the tea to be evaluated and the standard tea in the multiple quality dimensions is determined, along with the difference information corresponding to each target difference item, including: The quality descriptors and / or quality scores under the multiple quality dimensions, as well as the standard sample descriptors and / or standard sample scores, are compared to determine at least one target difference item between the tea to be evaluated and the standard tea in the multiple quality dimensions, and the difference information corresponding to each target difference item.

3. The method according to claim 2, characterized in that, The quality assessment information includes quality descriptive terms, and the quality standard sample information includes standard sample descriptive terms. The quality descriptive terms and standard sample descriptive terms under the multiple quality dimensions are compared to determine at least one target difference item between the tea to be evaluated and the standard sample tea in the multiple quality dimensions, as well as the difference information corresponding to each target difference item, including: The quality descriptors under the multiple quality dimensions are compared with the standard descriptors to determine the degree of matching between the quality descriptors under the multiple quality dimensions and the standard descriptors. If the degree of matching between the quality descriptor of the first quality dimension and the standard descriptor is less than the first threshold, then the first quality dimension is determined to be the target difference item, wherein the first quality dimension is any one of the multiple quality dimensions. The degree of matching between the quality descriptors of the first quality dimension and the standard sample descriptors is determined as the difference information corresponding to the target difference item.

4. The method according to claim 2, characterized in that, The quality assessment information includes a quality score, and the quality standard sample information includes a standard sample score. The quality scores under the multiple quality dimensions are compared with the standard sample scores to determine the target differences between the tea to be evaluated and the standard tea in the multiple quality dimensions, as well as the difference information corresponding to each target difference, including: The quality scores under the multiple quality dimensions are compared with the standard sample scores to determine the score difference between the quality descriptors under the multiple quality dimensions and the standard sample descriptors. If the difference between the quality score of the second quality dimension and the score of the standard sample is greater than the second threshold, then the second quality dimension is determined to be the target difference item, wherein the second quality dimension is any one of the multiple quality dimensions. The difference between the quality score of the second quality dimension and the score of the standard sample is determined as the difference information corresponding to the target difference item.

5. The method according to claim 1, characterized in that, The quality assessment information includes quality descriptive terms, and the difference dimension constraints include single-item dimension constraints; the step of analyzing at least one target difference item and the difference information corresponding to each target difference item based on the difference dimension constraints corresponding to the tea to be assessed, and determining the quality assessment result of the tea to be assessed, includes: If any quality descriptor of a target difference item contains a descriptor that matches the first keyword, then the quality assessment result of the tea to be evaluated is determined to be unsuccessful; or, If the degree of matching between the quality descriptor of any target difference item and the standard sample descriptor is less than the third threshold, then the quality assessment result of the tea to be evaluated is determined to be unsuccessful; or, If at least one of the target differences includes a key quality dimension, then the quality assessment result of the tea to be evaluated is determined to be unsuccessful, wherein the key quality dimension is determined based on the tea category of the tea to be evaluated.

6. The method according to claim 1, characterized in that, The difference dimension constraint includes a total dimension constraint, whereby the number of difference items exceeds a threshold. The step of analyzing at least one target difference item and the difference information corresponding to each target difference item based on the difference dimension constraint corresponding to the tea to be evaluated, and determining the quality evaluation result of the tea to be evaluated, includes: If the number of detected differences in the target difference item exceeds the quantity threshold, the quality assessment result of the tea to be evaluated is determined to be unsuccessful.

7. The method according to claim 1, characterized in that, After obtaining the quality assessment information of the tea to be evaluated in multiple quality dimensions, and the quality standard information of the standard tea in the multiple quality dimensions, the process further includes: Based on the quality assessment information of the tea to be evaluated in multiple quality dimensions, the comprehensive assessment information of the tea to be evaluated in multiple quality dimensions is determined; Obtain comprehensive standard sample information; Based on the comprehensive evaluation information and the comprehensive standard sample information, the comprehensive difference information between the tea to be evaluated and the standard tea is determined; Accordingly, the step of analyzing the at least one target difference item and the difference information corresponding to each target difference item based on the difference dimension constraints corresponding to the tea to be evaluated, and determining the quality evaluation result of the tea to be evaluated, further includes: Based on the difference dimension constraints and comprehensive difference constraints corresponding to the tea to be evaluated, the comprehensive difference information, the at least one target difference item, and the difference information corresponding to each target difference item are analyzed to determine the quality evaluation result of the tea to be evaluated.

8. The method according to claim 1, characterized in that, The process of obtaining quality assessment information for the tea to be evaluated across multiple quality dimensions includes: The tea evaluation interface displays evaluation controls for the multiple quality dimensions. The evaluation controls include preset quality information and / or input controls. The multiple quality dimensions and the preset quality information are configured based on the attributes of the tea to be evaluated. In response to an evaluation operation on an evaluation control under a target quality dimension, quality evaluation information under the target quality dimension is determined based on preset quality information selected by the evaluation operation and / or input quality information, wherein the target quality dimension is any one of the plurality of quality dimensions; In response to the evaluation completion event in the tea evaluation interface, quality evaluation information for each quality dimension is obtained.

9. A tea quality evaluation device, characterized in that, include: The acquisition module is configured to acquire quality assessment information of the tea to be evaluated in multiple quality dimensions, and quality standard information of the standard tea in the multiple quality dimensions, wherein the multiple quality dimensions are configured based on the attributes of the tea to be evaluated. The first determining module is configured to determine, based on the quality assessment information of the multiple quality dimensions and the corresponding quality standard sample information, at least one target difference item between the tea to be evaluated and the standard tea in the multiple quality dimensions, and the difference information corresponding to each target difference item. The second determining module is configured to analyze the at least one target difference item and the difference information corresponding to each target difference item based on the difference dimension constraint conditions corresponding to the tea to be evaluated, and determine the quality evaluation result of the tea to be evaluated. The difference dimension constraint conditions are configured based on the attributes of the tea to be evaluated and are used to indicate the constraint relationship between the difference item, the difference information and the quality evaluation result. The difference dimension constraint conditions include single dimension constraint conditions and / or total dimension constraint conditions.

10. A computing device, characterized in that, include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the tea quality evaluation method according to any one of claims 1-8.

11. A computer-readable storage medium, characterized in that, It stores computer-executable instructions that, when executed by a processor, implement the steps of the tea quality evaluation method according to any one of claims 1-8.

12. A computer program product, characterized in that, Includes a computer program / instruction that, when executed by a processor, implements the steps of the tea quality evaluation method according to any one of claims 1-8.

Images