A supplier dynamic window management method and system

Abstract

The present invention is applicable to the field of window management technology, and in particular to a supplier dynamic window management, the method comprising: determining at least two emotional categories, collecting vocabulary, and clustering the vocabulary into emotional categories, calculating the emotional intensity value of the vocabulary, and constructing an emotional dictionary; defining a reply sentence pattern, and establishing a link relationship between the reply sentence pattern and the emotional category; building a supplier's service list, adding key items, subscribing to message push in the service list, and obtaining emotional features. The present invention can sort message pushes by calculating the total intensity value of emotional features, so that important message pushes can be processed in a timely manner, which greatly improves the supplier's satisfaction and also improves the stability of the supply chain. By generating emotional responses, it is possible to better interact with suppliers in a personalized manner, improve the quality of dialogue, and also prevent excessive use of emotional replies, which may cause supplier disgust or distrust.

Description

Technical Field

[0001] The present invention relates to the technical field of window management, and in particular to a supplier dynamic window management method and system. Background Art

[0002] With the development of economic globalization, modern manufacturing enterprises no longer produce all the components of terminal products alone, but are more focused on establishing mature and reliable supply chains, integrating existing resources with the core components produced by the enterprise, and reducing production costs while ensuring the stability of the production process.

[0003] Due to the optimization and improvement of the industrial structure, modern manufacturing enterprises may have hundreds or thousands of suppliers, and suppliers will often change, so most suppliers have their own supply platforms; in the supply platform, the communication between suppliers is in the form of a chat window, and chat robots, intelligent assistants, etc. are used to reply to the window, but the emotional information in the window cannot be recognized, resulting in a poor user experience; for example, when some suppliers are eager to supply goods, or the windows of important suppliers may also be placed in the reply sequence, which greatly reduces the timeliness of the reply and affects the stability of the enterprise's supply chain; therefore, how to identify the emotional information in the window and give different responses is a technical problem that the present invention needs to solve. Summary of the invention

[0004] The purpose of the present invention is to provide a supplier dynamic window management method and system to solve the problem of "how to identify the emotional information in the window and give different responses" raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A supplier dynamic window management method, the method comprising:

[0007] Determine at least two sentiment categories, collect words, cluster the words into sentiment categories, calculate sentiment intensity values ​​of the words, and construct a sentiment dictionary;

[0008] Defining a reply sentence pattern, and establishing a link relationship between the reply sentence pattern and the emotion category;

[0009] Build a service list of suppliers and add key items, subscribe to the message push in the service list, obtain emotional features, and calculate the total intensity value of the emotional features, sort the message push in descending order of the total intensity value, grant pop-up window permissions to the message push and key items in the top column, and generate a normal response;

[0010] If the total strength value is greater than or equal to a preset threshold, the reply sentence pattern is locked, the reply sentence pattern and the common response are integrated to generate an emotional response, and the message push is processed using the emotional response;

[0011] If the total strength value is less than a preset threshold, the message push is directly processed using a normal response.

[0012] Furthermore, the steps of determining at least two emotion categories, collecting vocabulary, clustering the vocabulary into emotion categories, calculating the emotion intensity values ​​of the vocabulary, and constructing an emotion dictionary include:

[0013] Creating a sentiment vocabulary, segmenting the vocabulary into words, wherein the word is at least one;

[0014] Traversing the sentiment vocabulary table and calculating the sentiment intensity value of the vocabulary;

[0015] The vocabulary is added to the emotion category, all emotion categories are integrated, and an emotion dictionary is constructed.

[0016] Furthermore, the step of defining a reply sentence pattern and establishing a link relationship between the reply sentence pattern and the emotion category includes:

[0017] Select a template from a pre-established template library and introduce variables to determine the reply sentence pattern;

[0018] The semantic category of the reply sentence is determined, and based on the semantic category, a link relationship between the reply sentence and the emotion category is established.

[0019] Furthermore, the steps of building a service list of a supplier, adding key items, subscribing to message push in the service list, obtaining emotional features, and calculating the total intensity value of the emotional features include:

[0020] Receive message push from the service list, extract features from the message push, and obtain emotional features;

[0021] The emotional features are matched in the emotional dictionary, and the total intensity value of the emotional features is calculated.

[0022] Furthermore, the steps of sorting the message pushes in descending order of the total strength values, granting pop-up window permissions to the message pushes and key items at the top of the sort, and generating a normal response include:

[0023] Across the emotional features, a corresponding relationship between the total intensity value and the message push is established;

[0024] Using the total strength value as a keyword, sort the message push in descending order;

[0025] According to the sorting results, a preset number of messages are selected from the top rankings and pushed as important messages;

[0026] Grant pop-up permissions to the important information and key items.

[0027] Furthermore, the method further comprises:

[0028] Determine the difference between the total strength value and a preset threshold value, and if the total strength value is greater than or equal to the threshold value, locate the reply sentence pattern according to the emotional features in the message push;

[0029] The reply sentence pattern and the common response are integrated to generate an emotional response.

[0030] Furthermore, the method further comprises:

[0031] Create a neutral word category and collect neutral words;

[0032] It is determined whether the total strength value is zero. If it is zero, a neutral word is selected from the neutral word category, and the neutral word is filled into the reply sentence to generate a neutral response, and the message push is processed using the neutral response.

[0033] Furthermore, the system comprises:

[0034] A construction module, used to determine at least two sentiment categories, collect words, cluster the words into sentiment categories, calculate sentiment intensity values ​​of the words, and construct a sentiment dictionary;

[0035] A definition module, capable of defining a reply sentence pattern and establishing a link relationship between the reply sentence pattern and the emotion category;

[0036] The sorting module can build a service list of suppliers and add key items, subscribe to the message push in the service list, obtain emotional features, and calculate the total intensity value of the emotional features, sort the message push in descending order of the total intensity value, grant pop-up permissions to the message push and key items at the top of the sorting, and generate a normal response;

[0037] The judgment module can lock the reply sentence when the total strength value is greater than or equal to a preset threshold, merge the reply sentence and the ordinary response, generate an emotional response, and use the emotional response to process the message push; if the total strength value is less than the preset threshold, directly use the ordinary response to process the message push.

[0038] Furthermore, the definition module includes:

[0039] A creation unit, used for creating a sentiment vocabulary, dividing the vocabulary into words, wherein the word is at least one;

[0040] A traversal unit, capable of traversing the sentiment vocabulary table and calculating the sentiment intensity value of the vocabulary;

[0041] The hanging unit can hang the vocabulary to the emotion category, integrate all the emotion categories, and construct an emotion dictionary.

[0042] Furthermore, the definition module includes:

[0043] The selection unit can select a template from a pre-established template library, introduce variables, and determine the reply sentence pattern;

[0044] The establishing unit is used to determine the semantic category of the reply sentence pattern, and based on the semantic category, establish a link relationship between the reply sentence pattern and the emotion category.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] 1. By constructing a sentiment dictionary, the sentiment features in message push can be identified, so as to better understand the emotional state of suppliers. By calculating the total intensity value of the sentiment features, the message push can be sorted, so that important message push can be processed in a timely manner, which greatly improves the satisfaction of suppliers and also improves the stability of the supply chain. By generating emotional responses, better personalized interactions with suppliers can be carried out, which improves the quality of conversations. At the same time, it can also prevent the excessive use of emotional replies, which may cause supplier disgust or distrust.

[0047] 2. By judging the total intensity value of emotional features, we can better understand the emotional state of suppliers, and thus give different responses according to the emotional state of different suppliers. While ensuring the efficiency of the response, it also greatly improves the user experience and enhances the loyalty of suppliers. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 A flowchart of a supplier dynamic window management method provided by an embodiment of the present invention;

[0049] Figure 2 A first sub-flow chart of the supplier dynamic window management method provided by an embodiment of the present invention;

[0050] Figure 3 A second sub-flow chart of the supplier dynamic window management method provided by an embodiment of the present invention;

[0051] Figure 4 A third sub-flow chart of the supplier dynamic window management method provided by an embodiment of the present invention;

[0052] Figure 5A fourth sub-flow chart of the supplier dynamic window management method provided by an embodiment of the present invention;

[0053] Figure 6 A block diagram of the composition of the supplier dynamic window management system provided by an embodiment of the present invention;

[0054] Figure 7 A block diagram of the components of the building blocks in the supplier dynamic window management system provided by an embodiment of the present invention;

[0055] Figure 8 A block diagram of the components of the definition module in the supplier dynamic window management system provided by an embodiment of the present invention;

[0056] Fig. 9 A block diagram of the composition of a sorting module in a supplier dynamic window management system provided by an embodiment of the present invention;

[0057] Fig.10 A block diagram of the composition of the judgment module in the supplier dynamic window management system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0059] In Example 1, Figure 1 The implementation process of the supplier dynamic window management method provided by the embodiment of the present invention is shown and described in detail below:

[0060] S100: Determine at least two sentiment categories, collect words, cluster the words into sentiment categories, calculate sentiment intensity values ​​of the words, and construct a sentiment dictionary.

[0061] First, create multiple sentiment categories, such as positive, negative, neutral, and complex, and collect vocabulary. Then, classify the vocabulary into sentiment categories based on their meanings, assign different sentiment intensity values ​​to each vocabulary, and finally integrate all sentiment categories to construct a sentiment dictionary.

[0062] For example, construct three emotion categories, namely positive, negative and complex, and collect three words, namely happy, sad and contradictory. Then, transfer happy to the positive emotion category, sad to the negative emotion category, and contradictory to the complex emotion category. The emotion intensity value is divided into [0,5], with 4 points for happiness, 4 points for sadness, and 0 points for contradiction. Finally, integrate the emotion categories and construct an emotion dictionary.

[0063] S200: defining a reply sentence pattern, and establishing a link relationship between the reply sentence pattern and the emotion category.

[0064] In addition to the obvious emotional tendencies in vocabulary, sentence patterns also have emotional tendencies, so we define response sentences and establish the relationship between response sentences and emotional categories. For example, when you need to express sadness, the response sentence can be: "I feel very sad because..." By using response sentences, you can express emotions more accurately.

[0065] S300: Build a service list of suppliers and add key items, subscribe to the message push in the service list, obtain emotional features, and calculate the total intensity value of the emotional features, sort the message push in descending order of the total intensity value, grant pop-up permissions to the message push and key items at the top of the sort, and generate a normal response.

[0066] Build a supplier service platform, in which suppliers can communicate with each other. Each supplier has its own service list, similar to chat software and address book; select important or special suppliers from the service list as key items. For example, A's service list has suppliers B, C and D. Since the parts supplied by B are more important and there is no alternative supplier, B can be made the key item. Supplier C has higher requirements for communication efficiency and needs to reply to messages in time, so C can also be made the key item.

[0067] When a message push is received from a supplier in the service list (there may be multiple messages here, and they are mainly displayed in the form of windows), the emotional features in the message push, that is, the emotional words, are identified, and the total intensity value in the message push is calculated by querying the emotional dictionary. All the message push windows are sorted in order from large to small according to the total intensity value, and a normal response is generated. The normal response is the reply content, which can be obtained through the retrieval template.

[0068] Sort by total intensity value from large to small. The higher the ranking, the more eager the supplier is to express himself, and it needs to be processed in a timely manner.

[0069] S400: If the total strength value is greater than or equal to a preset threshold, the reply sentence is locked, the reply sentence and the ordinary response are integrated to generate an emotional response, and the message push is processed using the emotional response. If the total strength value is less than the preset threshold, the message push is directly processed using the ordinary response.

[0070] If the calculated total intensity is greater than or equal to the preset threshold, the reply sentence and ordinary content are integrated to generate an emotional response. The emotional response here is the reply content with emotional tendency. The emotional response is then used to reply to the message push, thereby realizing dynamic processing of supplier message push.

[0071] The preset threshold should be determined based on actual usage requirements. If the number of suppliers in the service list is greater, the preset threshold should be greater, so that all message push windows can be processed in a timely manner and processing efficiency can be improved.

[0072] When A receives the supply information windows of suppliers B, C and D at the same time, it is certainly impossible to display the three windows, which requires a sequence, so it is necessary to determine whether the total strength value in the supply information windows of suppliers B, C and D is greater than the preset threshold; if among suppliers B, C and D, the total strength value of C is greater than the preset threshold, it means that C may be in urgent need of communication at this time. The reason for the large total strength value of C may be the urgent need for goods or dissatisfaction with the previous batch of supplies, etc. At this time, it is necessary to respond emotionally to C, comfort him, stabilize C's emotions, and improve the quality of the conversation; and if the total strength value of B and D is less than the preset threshold, it means that B and D may be normal business exchanges. If an emotional response is made at this time, misunderstandings may occur. Excessive use of emotional responses may also make the responses appear exaggerated or inappropriate, causing disgust and distrust among suppliers.

[0073] In Example 2, Figure 2 The implementation process of the supplier dynamic window management method provided by an embodiment of the present invention is shown. The steps of determining at least two emotion categories, collecting words, clustering the words into emotion categories, calculating the emotion intensity values ​​of the words, and constructing an emotion dictionary are described in detail as follows:

[0074] S101: Create a sentiment vocabulary, and divide the vocabulary into words, where the word is at least one.

[0075] The collected vocabulary is divided into words. This has the advantage of being able to more accurately calculate the total intensity value of the sentiment feature. The sentiment vocabulary table is a table consisting of words and their sentiment intensity values. By querying the sentiment vocabulary table, the sentiment intensity value of the vocabulary can be calculated more quickly and accurately.

[0076] S102: traverse the sentiment vocabulary table to calculate the sentiment intensity value of the vocabulary.

[0077] The sentiment intensity value of a word is queried in the sentiment vocabulary table, and the sentiment intensity value of the word can be calculated by simply adding up the calculations.

[0078] S103: Add the vocabulary to the emotion category, integrate all emotion categories, and construct an emotion dictionary.

[0079] Classify words into different sentiment categories according to their meanings, and then integrate all sentiment categories to construct a sentiment dictionary.

[0080] In Example 3, Figure 3 The implementation process of the supplier dynamic window management method provided by the embodiment of the present invention is shown. The steps of defining the reply sentence pattern and establishing the link relationship between the reply sentence pattern and the emotional category are described in detail as follows:

[0081] S201: Select a template from a pre-established template library, introduce variables, and determine a reply sentence pattern.

[0082] Select a template library from existing data or public information, and introduce variables into the template according to actual usage needs, industries and suppliers, and then determine the reply sentence pattern based on the variables. The reply sentence pattern here is the reply template for the information push window;

[0083] For example, if you select a template such as "There are 300 shock-absorbing washers in stock. If needed, you can place an order by clicking Buy Now" from the template library, then use the shock-absorbing washers, 300 pieces, and Buy Now as variables and modify these variables according to different industries.

[0084] S202: Determine the semantic category of the reply sentence, and establish a link relationship between the reply sentence and the emotion category based on the semantic category.

[0085] By analyzing the reply sentence pattern, the semantic category is determined. The semantic category is the specific meaning it represents. For example, a reply sentence pattern can be used to urge delivery, while another reply sentence pattern can be used for daily communication. Then, based on the semantics of the reply sentence pattern, a link relationship between the reply sentence pattern and the emotional category is established. When a reply sentence pattern is needed, through this link relationship, appropriate words can be found in the corresponding emotional category, and the words can be filled into the reply sentence pattern to construct a reply sentence.

[0086] In Example 4, Figure 4 The implementation process of the supplier dynamic window management method provided by an embodiment of the present invention is shown. The steps of building a supplier's service list, adding key items, subscribing to message push in the service list, obtaining emotional features, and calculating the total intensity value of the emotional features are described in detail as follows:

[0087] S301: Receive a message push in a service list, extract features in the message push, and obtain emotional features.

[0088] Receive message push from suppliers in the service list, extract features from the window, and use the extracted features as sentiment features.

[0089] S302: Match the emotional features in the emotional dictionary and calculate the total intensity value of the emotional features.

[0090] By querying the sentiment vocabulary, the total intensity value of the sentiment characteristics is calculated. The larger the total intensity value is, the more urgent the supplier's communication needs are.

[0091] In Example 5, Figure 4 The implementation process of the supplier dynamic window management method provided by an embodiment of the present invention is shown. The steps of sorting the message pushes in descending order according to the total strength values, granting pop-up window permissions to the message pushes and key items in the top order, and generating a normal response are described in detail as follows:

[0092] S303: Across the emotional features, a corresponding relationship between the total intensity value and the message push is established.

[0093] There is a corresponding relationship between message push and emotional features, and there is a corresponding relationship between emotional features and total intensity values. That is to say, there is also a corresponding relationship between total intensity values ​​and message push. Crossing means through and utilizing emotional features.

[0094] S304: Using the total strength value as a keyword, sort the message push in descending order.

[0095] Sort the message push messages in descending order of total strength value.

[0096] S305: According to the sorting results, a preset number of messages are selected from the top rankings and pushed as important messages, and pop-up permissions are granted to important information and key items.

[0097] According to the sorting results, the top-ranked messages are identified as important messages, and the important messages are given pop-up window permissions. This is similar to chat software, where the windows for message push are folded together in order.

[0098] For example, when 5 message pushes are received at the same time, and the preset number is 1, the message pushes are first sorted according to the size of the total strength value, and the message push ranked first will automatically pop up and open, and the remaining 4 message pushes will continue to be collapsed in the order of total strength values. When the message push is opened, the first message push opened is the one with the highest total strength value among the remaining 4.

[0099] In Example 6, as shown in the attached specification Figure 5 As shown, different from Example 1, in this embodiment of the present invention, the method further includes:

[0100] S401: Determine the difference between the total strength value and a preset threshold value. If the total strength value is greater than or equal to the threshold value, locate a reply sentence pattern according to the emotional features in the message push;

[0101] S402: The reply sentence pattern and the common response are integrated to generate an emotional response.

[0102] Compare the total strength value with the preset threshold. If the total strength value is greater than or equal to the threshold, locate the reply sentence, integrate the reply sentence and the ordinary response, generate an emotional response, and use the emotional response to reply to the supplier's information push. If the total strength value is less than the threshold, directly use the ordinary response to reply to the message push.

[0103] In Example 7, different from Example 1, in this embodiment of the present invention, the method further includes:

[0104] Create a neutral word category and collect neutral words;

[0105] It is determined whether the total strength value is zero. If it is zero, a neutral word is selected from the neutral word category, and the neutral word is filled into the reply sentence to generate a neutral response, and the message push is processed using the neutral response.

[0106] In the sentiment dictionary, a neutral word category is created, and the collected neutral words are classified into it. It is determined whether the total strength value in the message push is zero. If it is zero, it means that the message push is a neutral message push. Neutral words are selected from the neutral word category, and the neutral words are filled into the reply sentence to generate a neutral response, and the message push is replied with the neutral response.

[0107] Figure 6 The following is a structural block diagram of a supplier dynamic window management system provided by an embodiment of the present invention. The supplier dynamic window management system 1 includes:

[0108] A construction module 11 is used to determine at least two sentiment categories, collect words, cluster the words into sentiment categories, calculate sentiment intensity values ​​of the words, and construct a sentiment dictionary;

[0109] A definition module 12, capable of defining a reply sentence pattern and establishing a link relationship between the reply sentence pattern and the emotion category;

[0110] The sorting module 13 can build a service list of suppliers, add key items, subscribe to the message push in the service list, obtain emotional features, and calculate the total strength value of the emotional features, sort the message push in descending order of the total strength value, grant pop-up permissions to the message push and key items at the top of the sorting, and generate a normal response;

[0111] The judgment module 14 can lock the reply sentence when the total strength value is greater than or equal to a preset threshold, merge the reply sentence and the ordinary response, generate an emotional response, and use the emotional response to process the message push; if the total strength value is less than the preset threshold, directly use the ordinary response to process the message push.

[0112] Figure 7 The following is a structural diagram of the supplier dynamic window management system provided by an embodiment of the present invention. The definition module 11 includes:

[0113] A creating unit 111, used for creating a sentiment vocabulary, dividing the vocabulary into words, wherein the word is at least one;

[0114] A traversal unit 112 is capable of traversing the sentiment vocabulary table and calculating the sentiment intensity value of the vocabulary;

[0115] The attaching unit 113 may attach the vocabulary to the emotion category, integrate all the emotion categories, and construct an emotion dictionary.

[0116] Figure 8 The following is a structural block diagram of the supplier dynamic window management system provided by an embodiment of the present invention. The definition module 12 includes:

[0117] The selection unit 121 can select a template from a pre-established template library and introduce variables to determine a reply sentence pattern;

[0118] The establishing unit 122 is used to determine the semantic category of the reply sentence pattern, and establish a link relationship between the reply sentence pattern and the emotion category based on the semantic category.

[0119] Fig. 9 The structural block diagram of the supplier dynamic window management system provided by the embodiment of the present invention is shown, and the sorting module 13 includes:

[0120] The extraction unit 131 can receive the message push in the service list, extract the features in the message push, and obtain the emotional features;

[0121] A calculation unit 132 is capable of matching the emotional features in the emotional dictionary and calculating a total intensity value of the emotional features;

[0122] A spanning unit 133, used to span the emotional features and establish a corresponding relationship between the total intensity value and the message push;

[0123] An arranging unit 134 is used to sort the message pushes in descending order using the total strength value as a keyword;

[0124] The granting unit 135 can select a preset number of messages from the top rankings and push them as important messages according to the sorting result, and grant pop-up window permissions to the important information and key items.

[0125] Fig.10 The structural block diagram of the supplier dynamic window management system provided by the embodiment of the present invention is shown, and the judgment module 14 includes:

[0126] The positioning unit 141 can determine the size of the total strength value and a preset threshold value, and if the total strength value is greater than or equal to the threshold value, locate the reply sentence according to the emotional features in the message push;

[0127] The fusion unit 142 can fuse the reply sentence and the common response to generate an emotional response.

[0128] The construction module 11 is mainly used to complete step S100, the definition module 12 is mainly used to complete step S200, the sorting module 13 is mainly used to complete step S300, and the judgment module 14 is mainly used to complete step S400;

[0129] The creation unit 111 is mainly used to complete step S101, the traversal unit 112 is mainly used to complete step S102, and the hanging unit 113 is mainly used to complete step S103;

[0130] The creation unit 121 is mainly used to complete step S201, and the establishment unit 122 is mainly used to complete step S202;

[0131] The extraction unit 131 is mainly used to complete step S301, the calculation unit 132 is mainly used to complete step S302, the spanning unit 133 is mainly used to complete step S303, the arrangement unit 134 is mainly used to complete step S304, and the assignment unit 135 is mainly used to complete step S305.

[0132] The positioning unit 141 is mainly used to complete step S201, and the fusion unit 142 is mainly used to complete step S2. The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0133] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

[0134] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A supplier dynamic window management method, characterized in that: The method comprises: Determine at least two sentiment categories, collect words, cluster the words into sentiment categories, calculate sentiment intensity values ​​of the words, and construct a sentiment dictionary; Defining a reply sentence pattern, and establishing a link relationship between the reply sentence pattern and the emotion category; Build a service list of suppliers and add key items, subscribe to the message push in the service list, obtain emotional features, and calculate the total intensity value of the emotional features, sort the message push in descending order of the total intensity value, grant pop-up window permissions to the message push and key items in the top column, and generate a normal response; If the total strength value is greater than or equal to a preset threshold, the reply sentence pattern is locked, the reply sentence pattern and the common response are integrated to generate an emotional response, and the message push is processed using the emotional response; If the total strength value is less than a preset threshold, directly use a normal response to process the message push; The step of defining a reply sentence pattern and establishing a link relationship between the reply sentence pattern and the emotion category comprises: Select a template from a pre-established template library and introduce variables to determine the reply sentence pattern; Determine the semantic category of the reply sentence, and establish a link relationship between the reply sentence and the sentiment category based on the semantic category; The steps of sorting the message pushes in descending order of the total strength values, granting pop-up window permissions to the message pushes and key items at the top of the sort, and generating a normal response include: Across the emotional features, a corresponding relationship between the total intensity value and the message push is established; Using the total strength value as a keyword, sort the message push in descending order; According to the sorting results, a preset number of messages are selected from the top rankings and pushed as important messages; Grant pop-up window permissions to the important messages and key items; The method further comprises: Create a neutral word category and collect neutral words; Determine whether the total strength value is zero. If it is zero, select a neutral word from the neutral word category, fill the neutral word into the reply sentence, generate a neutral response, and use the neutral response to process the message push; The method further comprises: Determine the magnitude of the total strength value and a preset threshold value, and if the total strength value is greater than or equal to the threshold value, locate a reply sentence pattern according to the emotional features in the message push; fusing the reply sentence pattern and the common response to generate an emotional response; The steps of determining at least two sentiment categories, collecting vocabulary, clustering the vocabulary into sentiment categories, calculating sentiment intensity values ​​of the vocabulary, and constructing a sentiment dictionary include: Creating a sentiment vocabulary, segmenting the vocabulary into words, wherein the word is at least one; Traversing the sentiment vocabulary table and calculating the sentiment intensity value of the vocabulary; The vocabulary is added to the emotion category, all emotion categories are integrated, and an emotion dictionary is constructed; The steps of building a service list of suppliers, adding key items, subscribing to message push in the service list, obtaining emotional features, and calculating the total intensity value of the emotional features include: Receive the message push in the service list, extract the features in the message push, and obtain the emotional features; The emotional features are matched in the emotional dictionary, and the total intensity value of the emotional features is calculated.

Images