An intelligent message processing method and system thereof

Abstract

The present application relates to an intelligent message processing method and system thereof, belonging to the field of logistics intelligent customer service. The method includes obtaining merchant feedback information; identifying keyword information; determining the current logistics status information; determining the current processing method information; analyzing the feedback object information; determining whether the feedback object information is merchant information; if so, processing the feedback object information according to the current logistics status information using the current processing method information; if not, determining the current logistics node information; sending the current processing method information to the network corresponding to the current logistics node information and feedbacking the processed information to the merchant corresponding to the merchant feedback information, improving the problem that e-commerce cannot reply to customers in time and has great manual limitations. The present application has the effect that through the order number and keyword information, the intelligent customer service gives a solution that meets the customer's intention, without the need for manual online to solve and answer customer questions, improving the intelligence and timeliness of the customer service.

Description

Technical Field

[0001] This application relates to the field of intelligent logistics customer service, and particularly to an intelligent message processing method and system thereof. Background Art

[0002] Express delivery, also known as express or express freight, refers to a new type of transportation method in which logistics enterprises (including freight forwarders) deliver the documents or packages entrusted by users from the sender to the recipient door-to-door (hand-to-hand) quickly and safely through their own independent networks or in a joint cooperation (i.e., networking) manner.

[0003] With the development of e-commerce platforms, the volume of express delivery business has increased explosively. E-commerce platforms often receive customers' logistics problems, and then transfer the feedback problems to express delivery companies. The express delivery companies give corresponding logistics status information and even take measures, such as returning, etc., to solve customers' problems and improve customers' satisfaction with goods and logistics.

[0004] Regarding the above related technologies, the inventor believes that the messages feedback from e-commerce to express delivery companies are often limited by large amounts of information and untimely responses. There may even be a situation where the logistics customer service can only reply the next day during the sleeping hours at night, resulting in e-commerce being unable to reply to customers in a timely manner, with significant manual limitations and room for improvement. Summary of the Invention

[0005] In order to improve the problem that e-commerce cannot reply to customers in a timely manner and has significant manual limitations, this application provides an intelligent message processing method and system thereof.

[0006] In a first aspect, this application provides an intelligent message processing method, adopting the following technical solutions:

[0007] An intelligent message processing method includes:

[0008] Obtain merchant feedback information;

[0009] Identify keyword information in the merchant feedback information, where the keyword information includes the current waybill number information and the current intention information;

[0010] Perform matching analysis based on the logistics status information stored in the preset logistics database and the current waybill number information to determine the logistics status corresponding to the current waybill number information, and define this logistics status as the current logistics status information;

[0011] Perform matching analysis based on the processing method information stored in the preset method database and the current intention information to determine the processing method corresponding to the current intention information, and define this processing method as the current processing method information;

[0012] Analyze the feedback object information in the current processing method information;

[0013] Determine whether the feedback object information is the preset merchant information;

[0014] If so, process the feedback object information according to the current processing method information based on the current logistics status information;

[0015] If not, perform a matching analysis based on the logistics node information stored in the preset node database and the current logistics status information to determine the logistics node in the current logistics status information, and define this logistics node as the current logistics node information;

[0016] Send the current processing method information to the network point corresponding to the current logistics node information and feedback the preset processed information to the merchant corresponding to the merchant feedback information.

[0017] By adopting the above technical solution, the status of the object is determined by identifying the order number, then the intention of the customer is analyzed based on the keyword information, and a solution that meets the customer's intention is given according to the object status and the customer's intention. And when an operation is required, the content that needs to be operated can be timely and real-time feedback to the logistics node. The logistics company does not need to solve and answer customer questions manually online, which improves the intelligence and timeliness of customer service.

[0018] Optionally, the method for identifying the current intention information in the keyword information of the merchant feedback information includes:

[0019] Analyze the merchant feedback information to determine the vocabulary information in the merchant feedback information;

[0020] Perform a matching analysis based on the logistics keyword information stored in the preset keyword database and the vocabulary information to determine the logistics keyword stored in the vocabulary information, and define this logistics keyword as the effective logistics keyword information;

[0021] Perform a matching analysis based on the logistics vocabulary information stored in the preset vocabulary database and the effective logistics keyword information to determine the logistics vocabulary corresponding to the effective logistics keyword information, and define this logistics vocabulary as the effective logistics vocabulary information;

[0022] Match the effective logistics vocabulary information with the vocabulary information one by one;

[0023] If the match is successful, select the successfully matched effective logistics vocabulary information and define it as the first effective logistics vocabulary information;

[0024] Perform a matching analysis based on the intention information stored in the preset intention database and the first effective logistics vocabulary information to determine the intention corresponding to the first effective logistics vocabulary information, and define this intention as the first effective intention information;

[0025] Output the first valid intent information as the current intent information;

[0026] If the match is unsuccessful, output the preset package tracking operation information as the current intent information.

[0027] By adopting the above technical solution, the corresponding keywords are identified from the vocabulary, and then matched with the logistics vocabulary with keywords in the logistics thesaurus. If the match is successful, the intent information is output normally. If it is not successful, it means that it is not a keyword or not related to logistics, so the package tracking operation information is directly used as the fallback intent for output, improving the intelligence of customer service feedback and the ability to handle exceptions.

[0028] Optionally, the method of matching the valid logistics vocabulary information with the vocabulary information one by one includes:

[0029] Screen out the character information in the valid logistics vocabulary information that is not the valid logistics keyword information, and define it as the valid character information;

[0030] Remove the valid character information from the valid logistics vocabulary information to obtain the new logistics vocabulary information;

[0031] Judge whether the new logistics vocabulary information is the logistics vocabulary information stored in the vocabulary database;

[0032] If so, output the new logistics vocabulary information and output the match success information;

[0033] If not, screen out the character information in the vocabulary information that is not the valid logistics keyword information, and define it as the invalid character information;

[0034] Perform matching analysis according to the correlation degree information, valid character information, and invalid character information stored in the preset correlation database to determine the correlation degree of the valid character information and the invalid character information, and define this correlation degree as the valid correlation degree information;

[0035] Judge whether the valid correlation degree information is greater than the preset correlation threshold information;

[0036] If so, output the valid logistics vocabulary information and output the match success information;

[0037] If not, output the match unsuccessful information.

[0038] By adopting the above technical solution, first determine whether the literal meaning is still clear and it is a logistics term after removing the invalid characters, so as to extract and combine the new terms to form real logistics terms, reducing the situation of missed recognition caused by the inclusion of tone and other words in the middle. For some words with the same pronunciation but different typed characters, by analyzing the correlation degree between the invalid character information and the valid character information, the real terms can be effectively recognized, further reducing the possibility of missed recognition and improving the ability and intelligence of keyword recognition.

[0039] Optionally, the method for outputting the current intention information includes:

[0040] Obtain the negative word information in the merchant feedback information;

[0041] Judge whether the negative word information exists;

[0042] If it exists, judge whether the position information of the negative word information is before the first valid logistics term information;

[0043] If so, perform matching analysis according to the intention information stored in the intention database, the first valid logistics term information, and the negative word information to determine the intention corresponding to the first valid logistics term information and the negative word information, and define this intention as the second valid intention information;

[0044] Output the second valid intention information as the current intention information;

[0045] If not, output the first valid intention information as the current intention information;

[0046] If it does not exist, output the first valid intention information as the current intention information.

[0047] By adopting the above technical solution, determine whether it is the positive intention of the keyword by judging whether there is a negative word in front of the keyword, reducing the situation of misjudgment caused by the existence of negative words and improving the accuracy of intention analysis.

[0048] Optionally, the method for outputting the second valid intention information as the current intention information includes:

[0049] Obtain the assumed word information in the merchant feedback information;

[0050] Judge whether the assumed word information exists;

[0051] If it exists, judge whether the position information of the assumed word information is before the negative word information;

[0052] If so, update the second valid intention information to the first valid intention information and output it as the current intention information;

[0053] If not, output the second valid intention information as the current intention information;

[0054] If it does not exist, output the second valid intention information as the current intention information.

[0055] By adopting the above technical solution, through the hypothetical vocabulary in front of the negative word, it is determined whether the vocabulary where the negative word exists is an affirmative statement, and it is determined whether the keyword is the state to be executed, and finally the true intention information is determined, further improving the accuracy of intention analysis.

[0056] Optionally, the method for further outputting the current intention information includes:

[0057] Obtain the quantity information of the current intention information in the merchant feedback information;

[0058] Judge whether the quantity information is greater than 1;

[0059] If it is equal to 1, output it with the current intention information;

[0060] If it is greater than 1, obtain the valid vocabulary number information;

[0061] Perform matching analysis according to the priority information and the valid vocabulary number information stored in the preset purpose database to determine the priority corresponding to the valid vocabulary number information, and define this priority as the current priority information;

[0062] Filter out the valid vocabulary number information with the largest current priority information and output the corresponding valid logistics vocabulary information.

[0063] By adopting the above technical solution, although there are multiple intention information corresponding to the keyword, analyzing its order to obtain the priority, so that the most priority one in the current intention can be quickly identified, that is, the most important purpose, improving the accuracy of intention recognition.

[0064] Optionally, the method for processing the feedback object information according to the current logistics status information using the current processing method information includes:

[0065] Perform matching analysis according to the possible process information stored in the preset process database and the current logistics status information to determine the possible process corresponding to the current logistics status information, and define this possible process as the current possible process information;

[0066] Match the current possible process information with the current processing method information one by one;

[0067] If the match is successful, output the current processing method information;

[0068] If the matching is unsuccessful, the parcel inquiry operation information is output as the current processing method information.

[0069] By adopting the above technical solution, it is possible to determine whether an operation can be performed by analyzing the processes and situations that can be carried out based on the current logistics status. When the time for processing has passed, the corresponding operation is no longer performed, improving the timeliness of intent processing.

[0070] Optionally, the method of outputting the parcel inquiry operation information as the current processing method information when both the effective logistics vocabulary information and the vocabulary information do not match includes:

[0071] Match the preset query vocabulary information with the vocabulary information one by one;

[0072] If the matching is successful, analyze the starting point information and the ending point information in the current order number information;

[0073] Perform matching analysis based on the node information and node time information stored in the preset node database, as well as the starting point information and the ending point information, to determine the node and node time corresponding to the starting point information and the ending point information. Define this node as the pre-node information and define this node time as the pre-node time information;

[0074] Obtain the segmented node information and segmented node time information of the current logistics status information;

[0075] Calculate the difference between the segmented node time information and the corresponding pre-node time information one by one, and define this difference as the actual difference information;

[0076] Judge whether the actual difference information falls within the corresponding time threshold information;

[0077] If it falls within, output the parcel inquiry operation information as the current processing method information;

[0078] If it does not fall within, filter out the segmented node information that does not fall within the time threshold information and obtain the abnormal cause information;

[0079] Perform matching analysis based on the solution information stored in the preset solution database and the abnormal cause information to determine the solution corresponding to the abnormal cause information, and define this solution as the current solution information;

[0080] Output the abnormal cause information and the current solution information;

[0081] If the matching is unsuccessful, output the parcel inquiry operation information as the current processing method information.

[0082] By adopting the above technical solution, by analyzing the abnormal states and abnormal possibilities in the logistics process, it enables the customer to quickly understand the customer's questions according to the actual situation and give corresponding reason explanations and processing solutions even though the customer does not explicitly express a certain problem, improving the intelligence and pertinence of the customer service.

[0083] Optionally, the method for calculating the actual difference information includes:

[0084] Obtain the historical vocabulary information corresponding to the current order number information;

[0085] Judge whether there is the preset return keyword information in the historical vocabulary information;

[0086] If there is return keyword information, obtain the node that receives the historical vocabulary information corresponding to the return keyword information, and define this node as the first new starting point information;

[0087] Perform matching analysis based on the node information, node time information, first new starting point information, and starting point information stored in the node database to determine the nodes and node times corresponding to the first new starting point information and starting point information, define this node as the first new node information, and define this node time as the first new node time information;

[0088] Update the first new node information to the pre-node information and recalculate the actual difference information;

[0089] If there is no return keyword information, judge whether there is the preset change keyword information in the historical vocabulary information;

[0090] If there is change keyword information, extract the first new ending point information from the historical vocabulary information and obtain the node that receives the historical vocabulary information corresponding to the change keyword information, and define this node as the second new starting point information;

[0091] Perform matching analysis based on the node information, node time information, second new starting point information, and first new ending point information stored in the node database to determine the nodes and node times corresponding to the second new starting point information and first new ending point information, define this node as the second new node information, and define this node time as the second new node time information;

[0092] Update the second new node information to the pre-node information and recalculate the actual difference information;

[0093] If there is no change keyword information, directly calculate the actual difference information.

[0094] By adopting the above technical solution, the historical vocabulary information corresponding to the current order number information is obtained and analyzed to determine whether it will cause changes to the logistics nodes, thereby determining the true logistics node information, preventing misjudgment, and improving the accuracy of obtaining abnormal states and abnormal possibilities in the logistics process.

[0095] In a second aspect, the present application provides an intelligent message processing system, adopting the following technical solution:

[0096] An intelligent message processing system includes:

[0097] An information acquisition module for acquiring merchant feedback information;

[0098] A processing module, connected to the information acquisition module and the judgment module, for storing and processing information;

[0099] An identification module, connected to the processing module, for identifying keyword information in the merchant feedback information, where the keyword information includes the current order number information and the current intention information;

[0100] The processing module performs matching analysis based on the logistics status information stored in the preset logistics database and the current order number information to determine the logistics status corresponding to the current order number information, and defines this logistics status as the current logistics status information;

[0101] The processing module performs matching analysis based on the processing method information stored in the preset method database and the current intention information to determine the processing method corresponding to the current intention information, and defines this processing method as the current processing method information;

[0102] An analysis module, connected to the processing module, for analyzing the feedback object information in the current processing method information;

[0103] A judgment module for judging whether the feedback object information is the preset merchant information;

[0104] If the judgment module determines that it is, the processing module processes the feedback object information according to the current logistics status information using the current processing method information;

[0105] If the judgment module determines that it is not, the processing module performs matching analysis based on the logistics node information stored in the preset node database and the current logistics status information to determine the logistics node in the current logistics status information, and defines this logistics node as the current logistics node information;

[0106] The processing module sends the current processing method information to the network point corresponding to the current logistics node information and feeds back the preset processed information to the merchant corresponding to the merchant feedback information.

[0107] By adopting the above technical solution, the object status is determined by identifying the single number, then the customer's intention is analyzed based on the keyword information, and a solution that meets the customer's intention is given according to the object status and the customer's intention. When an operation is required, the content to be operated can be timely and real-time fed back to the logistics node. The logistics company does not need to solve and answer customer questions manually online, which improves the intelligence and timeliness of the customer service.

[0108] In summary, the present application includes at least one of the following beneficial technical effects:

[0109] 1. Through the single number and keyword information, the intelligent customer service gives a solution that meets the customer's intention, without the need to solve and answer customer questions manually online, which improves the intelligence and timeliness of the customer service;

[0110] 2. By analyzing the current logistics status, when the time for processing has passed, the corresponding operation is no longer performed, which improves the timeliness of intention processing;

[0111] 3. By analyzing the abnormal status and abnormal possibility in the logistics process, quickly understand the customer's questions and give corresponding cause explanations and treatment solutions, which improves the intelligence and pertinence of the customer service. Description of the Drawings

[0112] Figure 1 It is a flowchart of an intelligent message processing method in an embodiment of the present application.

[0113] Figure 2 It is a flowchart of a method for identifying the current intention information in the keyword information of the merchant feedback information in an embodiment of the present application.

[0114] Figure 3 It is a flowchart of a method for matching the effective logistics vocabulary information with the vocabulary information one by one in an embodiment of the present application.

[0115] Figure 4 It is a flowchart of a method for outputting the current intention information in an embodiment of the present application.

[0116] Figure 5 It is a flowchart of a method for outputting the second effective intention information as the current intention information in an embodiment of the present application.

[0117] Figure 6 It is a flowchart of a method for further outputting the current intention information in an embodiment of the present application.

[0118] Figure 7 It is a flowchart of a method for processing the feedback object information according to the current processing method information based on the current logistics status information in an embodiment of the present application.

[0119] Figure 8 This is a flowchart for outputting the package tracking operation information as the current processing method information when the matching of valid logistics vocabulary information and vocabulary information in the embodiments of the present application fails.

[0120] Figure 9 This is a flowchart for calculating the actual difference information in the embodiments of the present application.

[0121] Figure 10 This is a module diagram of an intelligent message processing method in the embodiments of the present application. Detailed implementation manners

[0122] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application will be further described in detail below with reference to the Figures 1-10 accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0123] The embodiments of the present invention will be further described in detail below with reference to the drawings of the specification.

[0124] Refer to Figure 1 , the embodiments of the present invention provide an intelligent message processing method, and the main process of an intelligent message processing method is described as follows:

[0125] Step 100: Obtain merchant feedback information.

[0126] The merchant feedback information is the information that the merchant sorts out after receiving the customer's feedback content, which is a bit confused and has requirements for a certain product. The obtaining method is a unified collection method. For example, all merchants cooperating with this logistics company are added to a QQ group or a WeChat group, and then the merchants immediately send the corresponding questions in the group. It should be noted that since it is a merchant, the general feedback information will be sent after being processed and is relatively standardized.

[0127] Step 101: Identify the keyword information in the merchant feedback information, where the keyword information includes the current waybill number information and the current intention information.

[0128] The keyword information is the information of the vocabulary that can be recognized by the robot in the merchant feedback information. The current waybill number information is the waybill number in the keyword that can reflect the express information and status, which is the unique identification code of the express package and contains the detailed information of the recipient's express. It is usually composed of numbers and letters. The express waybill number can facilitate the express company, the sender and the recipient to track the express information in real time. The current intention information is the information of the operation that the customer wants to perform on the express with the current waybill number information. The identification method is character recognition, that is, the characters are matched with the items in the template, and a selection statement is used.

[0129] When the intention cannot be recognized or the user changes the intention statement, the items in the template will be trained and added to improve the scope and accuracy of intention recognition.

[0130] Step 102: Perform a matching analysis based on the logistics status information stored in the preset logistics database and the current order number information to determine the logistics status corresponding to the current order number information, and define this logistics status as the current logistics status information.

[0131] The current logistics status information is the information about the status of the commodity item corresponding to the current order number in the logistics. For example: the to-be-shipped status. The logistics database stores the mapping relationship between the logistics status information and the current order number information. It is obtained by real-time updating of the logistics system, that is, the logistics company will real-time update the logistics status of the shipped goods and the arrival nodes when reaching each node, and then input them into the logistics database for storage and recording. Users can view the past time nodes and status through the logistics database. When the system receives the current order number information, it automatically receives the corresponding real-time updated logistics status and outputs it as the current logistics status information.

[0132] Step 103: Perform a matching analysis based on the processing method information stored in the preset method database and the current intention information to determine the processing method corresponding to the current intention information, and define this processing method as the current processing method information.

[0133] The current processing method information is the information about the processing method adopted to deal with the intention in the current intention information. For example, if the customer's intention is to query, the corresponding processing method is to send out the corresponding materials. Another example is that if the customer's intention is to return the item, the corresponding processing method is to return the commodity for processing. The database stores the mapping relationship between the processing method information and the current intention information. It is input by the staff in this field in combination with the actual scenarios and experience, and will continue to be trained and added in the subsequent actual application process to enrich the different types of the current intention information and the corresponding processing methods. When the system receives the current intention information, it automatically finds the corresponding processing method and outputs it in the form of the current processing method information.

[0134] Step 104: Analyze the feedback object information in the current processing method information.

[0135] The feedback object information is the information about the object that the processing method is aimed at. The current processing method information also contains the corresponding object. For example, the corresponding object for querying is the merchant; the object for returning the item is the corresponding network point. The essence of the analysis method is a process of extraction, that is, extracting the corresponding object method.

[0136] Step 105: Determine whether the feedback object information is the preset merchant information.

[0137] The merchant information corresponding to the feedback message is the information of the merchant. The purpose of the judgment is to determine the requirement target.

[0138] Step 1051: If so, process the feedback object information according to the current processing method information based on the current logistics status information.

[0139] When it is yes, it means that the feedback object information is the merchant. Since it is a message sent by a merchant, the corresponding method in the current processing method information can be directly adopted for the corresponding object.

[0140] Step 1052: If not, perform matching analysis based on the logistics node information stored in the preset node database and the current logistics status information to determine the logistics node in the current logistics status information, and define this logistics node as the current logistics node information.

[0141] The current logistics node information is the information of the node of the commodity in the current logistics status information, which is essentially the node of the logistics where the commodity is located when a customer problem occurs at that time. The node database stores the mapping relationship between the logistics node information and the current logistics status information. The logistics status in the current logistics status information is extracted by the staff in this field, and then the corresponding nodes of the logistics network points are matched to obtain. It is essentially the node corresponding to the current logistics status information (i.e., the last updated piece of information in the logistics flow chart) in the logistics system. When the system receives the current logistics status information, it automatically finds the corresponding logistics node information from the database.

[0142] Step 106: Send the current processing method information to the network point corresponding to the current logistics node information and feedback the preset processed information to the merchant corresponding to the merchant feedback information.

[0143] When the execution object of the processing method is not the merchant, it means it is on the logistics company side. Then, when the current logistics node information is known, the current processing method information is automatically sent to the corresponding network point, and then a notice of having been processed is feedback to the merchant to reply to the merchant's question.

[0144] Refer to Figure 2 , the method for identifying the current intention information in the keyword information of the merchant feedback information includes:

[0145] Step 200: Analyze the merchant feedback information to determine the vocabulary information in the merchant feedback information.

[0146] Lexical information is information about words with semantics in the merchant feedback information. It can be a single character or a word composed of multiple characters connected together. This word needs to be found in the Xinhua Dictionary. The analysis process is a process of extraction, that is, searching character by character. When a word can be formed and the next character cannot form a word when added, the previous word is automatically extracted and the extraction continues from the next character and the characters following it in sequence.

[0147] Step 201: Perform a matching analysis based on the logistics keyword information and lexical information stored in the preset keyword database to determine the logistics keywords stored in the lexical information, and define this logistics keyword as the effective logistics keyword information.

[0148] Effective logistics keyword information is information about the key words that reflect a certain process in the logistics process. For example: "return", "check", etc. The keyword database stores the mapping relationship between logistics keyword information and lexical information. Substantially, logistics keywords are stored in the database, and then each character in the lexical information is matched with the keywords in the database one by one. When the two are the same, the mapping relationship between the two is recorded and stored. On the one hand, it can be accurately queried; on the other hand, a training model can be formed so that when the same word appears in the future, the corresponding logistics keyword can be quickly found. When the system receives lexical information, it automatically searches the database for the corresponding logistics keyword and outputs it.

[0149] Step 202: Perform a matching analysis based on the logistics vocabulary information and effective logistics keyword information stored in the preset vocabulary database to determine the logistics vocabulary corresponding to the effective logistics keyword information, and define this logistics vocabulary as the effective logistics vocabulary information.

[0150] Effective logistics vocabulary information is information about the logistics-related vocabulary that can be formed by the effective logistics keyword information. For example, the corresponding vocabulary for "urge" includes "urge the package, urge, give a reminder, hurry up and urge", etc. The database stores the mapping relationship between logistics vocabulary information and effective logistics keyword information, which is recorded and stored by the staff in this field based on long-term work experience and consulting relevant materials, etc. When the system receives the corresponding effective logistics keyword information, it automatically searches the database for the corresponding logistics vocabulary and outputs it as the effective logistics vocabulary information.

[0151] Step 203: Match the effective logistics vocabulary information with the lexical information one by one.

[0152] The purpose of the matching is to determine whether the vocabulary is consistent with the logistics vocabulary information in the template so as to clarify the intention of the vocabulary.

[0153] Step 2031: If the match is successful, select the effective logistics vocabulary information that matches successfully and define it as the first effective logistics vocabulary information.

[0154] The first valid logistics vocabulary information is the information of the logistics vocabulary that successfully matches the vocabulary information. If the match is successful, it means that the vocabulary information is exactly the same as the logistics vocabulary in the database.

[0155] Step 2032: If the match is unsuccessful, the preset package tracking operation information is output as the current intention information.

[0156] The package tracking operation information is the information of the operation to query the goods corresponding to the current order number information and the logistics status of the goods. If the match is unsuccessful, it means that it is not an intention in the database and the customer's intention cannot be accurately recognized. Then a fallback process is performed, and the package tracking operation information is output as the intention to indicate that the logistics company has received the message, so as to prevent the situation where the merchant does not take measures due to not receiving feedback information for a long time.

[0157] Step 204: Perform a matching analysis based on the intention information stored in the preset intention database and the first valid logistics vocabulary information to determine the intention corresponding to the first valid logistics vocabulary information, and define this intention as the first valid intention information.

[0158] The first valid intention information is the information of the intention corresponding to the logistics vocabulary corresponding to the first valid logistics vocabulary information. When it is a professional logistics vocabulary, it can also be converted into a logistics intention. There is a mapping relationship between the intention information and the first valid logistics vocabulary information stored in the database, and the staff in this field automatically inputs its meaning according to its logistics intention when inputting the first valid logistics vocabulary information according to the actual situation. When the system receives the first valid logistics vocabulary information with a successful match, it automatically finds the corresponding intention from the database and outputs it as the first valid intention information.

[0159] Step 205: Output the first valid intention information as the current intention information.

[0160] When the vocabulary matches the professional logistics vocabulary in the database and its true purpose is also the same as the first valid intention information, the customer's vocabulary is converted into a professional logistics vocabulary and a logistics intention is output.

[0161] Refer to Figure 3 , the method of matching the valid logistics vocabulary information with the vocabulary information one by one includes:

[0162] Step 300: Screen out the character information in the valid logistics vocabulary information that is not valid logistics keyword information, and define it as valid character information.

[0163] Valid character information refers to other characters in the valid logistics vocabulary information that are not valid logistics keyword information. The screening method is to retain the valid logistics keywords and select the rest. For example, "促一下", where the valid logistics keyword information is "促", then "一下" is the valid character information. It should be noted that although a valid logistics keyword information can find a valid logistics vocabulary information, there is not only one valid logistics keyword information in a valid logistics vocabulary information, and all valid logistics keyword information needs to be screened out.

[0164] Step 301: Eliminate valid character information from valid logistics vocabulary information to obtain new logistics vocabulary information.

[0165] The new logistics vocabulary information is the vocabulary information consisting of the remaining characters after removing other characters that are not valid logistics keyword information from the valid logistics vocabulary information.

[0166] Step 302: Determine whether the new logistics vocabulary information is the logistics vocabulary information stored in the vocabulary database.

[0167] The purpose of the judgment is to determine whether the complete meaning can still be expressed after removal.

[0168] Step 3021: If yes, output new logistics vocabulary information and output matching success information.

[0169] If so, it means that the complete meaning can still be expressed, and the match can still be successful when the valid character information is any word, avoiding the situation where the merchant makes a typo and cannot be recognized.

[0170] Step 3022: If not, the character information in the vocabulary information that is not valid logistics keyword information is filtered out and defined as non-valid character information.

[0171] Invalid character information refers to character information in the vocabulary information that is not valid logistics keyword information. If not, it means that after removing the valid characters, a valid logistics word cannot be expressed, and it is still necessary to combine the valid character information. At this time, it is necessary to analyze whether there are homophones in the vocabulary information.

[0172] Step 303: performing matching analysis based on the association degree information and the valid character information and the invalid character information stored in the preset association database to determine the association degree between the valid character information and the invalid character information, and defining the association degree as the valid association degree information.

[0173] The effective association degree information is the association degree between effective character information and non-effective character information. The mapping relationship of effective character information, non-effective character information, and association degree information is stored in the database. Essentially, when effective character information and non-effective character information are input, the output is the information of the association degree between the two. This is for the staff in the field to analyze all the effective character information and then input all the corresponding relevant non-effective character information. It should be noted that when the non-effective character is not in the database, it is default defined as having the worst association degree. When the system receives effective character information and non-effective character information, it automatically searches for the corresponding association degree in the database and then outputs it as the effective association degree information.

[0174] Step 304: Determine whether the effective association degree information is greater than the preset association threshold information.

[0175] The association threshold information is the threshold information of the association degree. In this system, in order to express the corresponding association degree, the association degree is represented by numbers. If it is greater than the association threshold, it means that the two are related, which may be homophonic words or misspelled words; if it is less than, it means that they are not related at all and are completely two different meanings, etc.

[0176] Step 3041: If so, output the effective logistics vocabulary information and output the matching success information.

[0177] If so, it means that they are associated at this time, which means that although the characters are different, they are completely related. Therefore, in the absence of other semantics, the effective logistics vocabulary information can be output and the matching success information can be output. It should be noted here that this judgment is made when the effective logistics vocabulary information and the vocabulary information are somewhat inconsistent. If they are consistent, it can be directly output. If they are inconsistent, the method of steps 300 - 3041 is used for fuzzy recognition.

[0178] Step 3042: If not, output the matching failure information.

[0179] If so, it means that even fuzzy recognition cannot recognize it at this time, so it is not a logistics vocabulary at all, and the matching failure information can be output.

[0180] Refer to Figure 4 , the method for outputting the current intention information includes:

[0181] Step 400: Obtain the negative word information in the merchant feedback information.

[0182] The negative word information is the information of negative words. The obtaining method is to match all the negative words with the merchant feedback information, and if the match is successful, the corresponding words are screened out.

[0183] Step 401: Determine whether negative term information exists.

[0184] Step 4011: If it exists, determine whether the position information of the negative term information is before the first valid logistics term information.

[0185] The position information is the position where the term information is located, which refers to the position of the negative term information here. If it exists, it means that there is a negative term at this time, and it is possible that the intention corresponding to the keyword information is not a positive intention but a reverse intention. Therefore, it is necessary to determine whether this negative term appears before the first valid logistics term information. It should be noted here that the meaning of "before" is the position immediately before, and there are no intervening characters in the middle.

[0186] Step 4012: If it does not exist, output the first valid intention information as the current intention information.

[0187] If it does not exist, it means that there is no negative term, indicating that the intention at this time is still the intention of the term corresponding to the first valid logistics term information. Therefore, the first valid intention information can be output as the current intention information.

[0188] Step 4021: If so, perform matching analysis based on the intention information stored in the intention database, the first valid logistics term information, and the negative term information to determine the intention corresponding to the first valid logistics term information and the negative term information, and define this intention as the second valid intention information.

[0189] The second valid intention information is the information of the intention generated by the combination of the first valid logistics term information and the negative term information. The intention database also stores the mapping relationship between the intention information, the first valid logistics term information, and the negative term information. Among them, the first valid logistics term information and the negative term information are combined to form a reverse logistics term, and this term is also recorded in the database. Therefore, when the first valid logistics term information and the negative term information are input, the system automatically searches the database for the corresponding intention and outputs it as the second valid intention information. If so, it means that the new term formed by the negative term and the first valid logistics term information at this time is the reverse term of the first valid logistics term information, such as "not return".

[0190] Step 4022: If not, output the first valid intention information as the current intention information.

[0191] If not, it means that the negative term information is after or there are other characters in the middle although it is before. There may be other meanings, but in most cases, they do not affect each other. Therefore, the first valid intention information is still output as the current intention information.

[0192] Step 403: Output the second valid intention information as the current intention information.

[0193] Refer to Figure 5 , the method of outputting the second valid intention information as the current intention information includes:

[0194] Step 500: Obtain the assumed word information in the merchant feedback information.

[0195] The assumed word information is a word meaning conjecture, prediction, or the existence of such a situation, such as: "if". The obtaining method is to match all the assumed words with the merchant feedback information, and if the match is successful, the corresponding words are filtered out.

[0196] Step 501: Determine whether the assumed word information exists.

[0197] Step 5011: If it exists, determine whether the position information where the assumed word information is located is before the negative word information.

[0198] The position information is the position where the word information is located, here referring to the position of the assumed word information. If it exists, it means that there is an assumed word at this time, then the intention corresponding to the possible keyword information may not be a positive intention or a negative intention, but what means to be taken if such a situation occurs, such as "if returned", which means not to return, but a questioning tone, so it is necessary to determine whether this assumed word information appears before the first valid logistics word information. Here, it should be noted that the meaning of "before" here is the position immediately before, without any intervening characters in the middle.

[0199] Step 5012: If it does not exist, output the second valid intention information as the current intention information.

[0200] Among them, it is explained here that there is no assumed word, indicating that the meaning of the keyword is still the same at this time. If there is a negative word, it is the intention of the word corresponding to the second valid intention information, then the second valid intention information can be output as the current intention information; of course, if there is no negative word, the first valid intention information is output as the current intention information.

[0201] Step 5021: If so, update the second valid intention information to the first valid intention information and output it as the current intention information.

[0202] If so, it means that there is an assumed word and it is before the negative word and the keyword, and the assumed word generally is not the real intention. For example, "if not rush the order" means not the intention of not rushing the order, but the intention of rushing the order, then the first valid intention information is still output as the current intention information.

[0203] Step 5022: If not, then output the second valid intention information as the current intention information.

[0204] If not, it means that although there are assumed words, they are not related to the newly formed reverse logistics vocabulary. Therefore, the second valid intention information is still output as the current intention information.

[0205] Refer to Figure 6 , the method for further outputting the current intention information includes:

[0206] Step 600: Obtain the quantity information of the current intention information in the merchant feedback information.

[0207] The quantity information is the information of the number of the current intention information.

[0208] Step 601: Determine whether the quantity information is greater than 1.

[0209] The purpose of the determination is to determine whether there are multiple intentions in the feedback information that interfere with the robot's recognition.

[0210] Step 6011: If it is equal to 1, then output the current intention information.

[0211] If it is equal to 1, it means that the intention is clear and unique, so it can be directly output.

[0212] Step 6012: If it is greater than 1, then obtain the valid vocabulary number information.

[0213] The valid vocabulary number information is the information of the number of the vocabulary. The purpose of obtaining it is to determine the information of all other contents of the vocabulary and make distinctions. If it is greater than 1, then obtain the corresponding number information. In the database, not only the valid vocabulary number information is stored, but also the corresponding numbers are stored.

[0214] Step 602: Perform matching analysis according to the priority information and the valid vocabulary number information stored in the preset target database to determine the priority corresponding to the valid vocabulary number information, and define this priority as the current priority information.

[0215] The current priority information is the information of the number of the valid logistics vocabulary information corresponding to the valid vocabulary number information. The database stores the mapping relationship between the priority information and the valid vocabulary number information. When the staff in this field input the logistics vocabulary into the database, they also input the priority information according to the logistics process. For example: the priority of delivery is greater than that of signing for receipt, the priority of reminder is greater than that of signing for receipt, the priority of address change is greater than that of signing for receipt, etc. When the system receives the valid vocabulary number information, it automatically finds the corresponding priority from the database and outputs it as the current priority information.

[0216] Step 603: Screen out the valid vocabulary number information with the largest current priority information and output the corresponding valid logistics vocabulary information.

[0217] The purpose of the screening here is to clarify that when there are multiple intentions in this statement, screen out the most important purpose, that is, the most fundamental purpose for output. For example, in the statement "Hurry up and remind, otherwise the user will return it", there are two intentions in this statement, "remind about the order" and "return the goods". Since the priority of reminding about the order is greater than the priority of returning the goods, the actual intention is "remind about the order", so the purpose of reminding about the order can be output. And when it comes to "Don't remind about the order anymore, I want to return the goods", the intentions at this time are "don't remind about the order" and "return the goods". At this time, the priority of returning the goods is greater than the priority of not reminding about the order, so the actual intention is "return the goods".

[0218] Refer to Figure 7 , the method of processing the feedback object information according to the current logistics status information using the current processing method information includes:

[0219] Step 700: Perform a matching analysis based on the possible process information stored in the preset process database and the current logistics status information to determine the possible process corresponding to the current logistics status information, and define this possible process as the current possible process information.

[0220] The current possible process information is the information about the operations that the merchant can take on the goods under the current logistics status information. For example: if the current logistics status information is that the goods have reached a certain network point, the possible operations are to remind about the order, return, change the itinerary, query, etc., but it is not possible to remind about the delivery, etc. Or if the current logistics status information is that the goods have been signed for, the possible operations are to return the goods, query, etc., and it is not possible to change the itinerary or remind about the order. The mapping relationship between the possible process information and the current logistics status information is stored in the database and is recorded and input by the staff in this field according to the logistics regulations and the experience in real life. When the system receives the current logistics status information, it automatically searches the database for the corresponding possible process and outputs it as the current possible process information.

[0221] Step 701: Match the current possible process information with the current processing method information one by one.

[0222] The matching method can be literal matching of the two, or synonym matching.

[0223] Step 7011: If the match is successful, output the current processing method information.

[0224] If the match is successful, it means that the method corresponding to the current processing method information can be adopted, and the processing is carried out according to the current processing method information.

[0225] Step 7012: If the match is unsuccessful, the parcel tracking operation information is output as the current processing method information.

[0226] If the match is unsuccessful, it means that the timeliness of the current processing method information has passed and no further operations can be performed. To indicate that the customer service has read it, the parcel tracking operation information is output as the current processing method information.

[0227] Refer to Figure 8 , if both the effective logistics vocabulary information and the vocabulary information do not match successfully, the method of outputting the parcel tracking operation information as the current processing method information includes:

[0228] Step 800: Match the preset question vocabulary information with the vocabulary information one by one.

[0229] The question vocabulary information is information where the user only expresses a question but there is no specific keyword, such as "What's going on". At this time, the user's intention is not clear and it cannot be matched with the effective logistics vocabulary information.

[0230] Step 8001: If the match is successful, analyze the starting point information and the ending point information in the current order number information.

[0231] If the match is successful, it means that the merchant has expressed a question at this time, indicating that the merchant has some doubts about the order number. If it is a doubt about the product, the merchant usually answers it actively by himself. Therefore, the doubts sent by the merchant in the group are generally about logistics, so the abnormal situations need to be analyzed.

[0232] Step 8002: If the match is unsuccessful, the parcel tracking operation information is output as the current processing method information.

[0233] If the match is unsuccessful, it means that at this time it does not match the effective logistics vocabulary information nor the question vocabulary information, indicating that the intention cannot be recognized at this time, and the parcel tracking operation information is normally output as the current intention information.

[0234] Step 801: Match and analyze according to the node information, node time information, starting point information, and ending point information stored in the preset node database to determine the node and node time corresponding to the starting point information and the ending point information. Define this node as the pre-node information and this node time as the pre-node time information.

[0235] The pre-node information is the information of the network points that are theoretically passed through during the process from the starting point to the ending point and the processes at these network points, including the information of the network points and processes corresponding to the starting point information and the ending point information. The pre-node time information is the information of the time that theoretically stays at the pre-node information. In the database, there is a mapping relationship of the starting point information, the ending point information, the node information, and the node time information, which is automatically established by the logistics system. Of course, in the early stage, it is also formed by those skilled in the art through simulation analysis according to reasonable distributions, reasonable paths in real life, and existing logistics routes. When the environment is updated, the database is also updated in real time. When the system receives the starting point information and the ending point information, it automatically searches in the database for the corresponding nodes and node times, and outputs them as pre-node information and pre-node time information respectively.

[0236] Step 802: Obtain the segmented node information and the segmented node time information of the current logistics status information.

[0237] The segmented node information is the information of the nodes actually passed through under the current logistics status information. The segmented node time information is the information of the time actually stayed at the segmented node information. The obtaining method is to input the nodes and node times simultaneously when the current logistics status information is updated, and then extract them.

[0238] Step 803: Calculate the difference between the segmented node time information and the corresponding pre-node time information one by one, and define this difference as the actual difference information.

[0239] The actual difference information is the information of the difference between the segmented node time information and the corresponding pre-node time information. The calculation method is to calculate one by one correspondence, that is, when the segmented node information and the pre-node information match, subtract the pre-node time information that matches successfully from the segmented node time information to obtain the actual difference information.

[0240] Step 804: Determine whether the actual difference information falls within the corresponding time threshold information.

[0241] The time threshold information is the information of the time delay allowed to exist. That is, the time delay corresponding to this time threshold is allowed.

[0242] Step 8041: If it falls within, output the package tracking operation information as the current processing method information.

[0243] If it falls within, it means that there is no problem in the logistics process, and the question information of the customer feedback by the merchant is not a problem corresponding to the logistics. Then, only the package tracking operation information can be output as the current processing method information.

[0244] Step 8042: If it does not fall within, screen out the segmented node information that does not fall within the time threshold information and obtain the abnormal cause information.

[0245] The abnormal reason information is the information about the abnormal reason for staying too long at the outlet corresponding to the segment node information. Generally, in the logistics industry, when the stay at the outlet is too long, the staff of the outlet will enter the corresponding reason into the system for easy reference and processing. The acquisition method is manual input. If it does not fall into the system, it means that the outlet corresponding to the segment node information is abnormal, and the abnormal reason needs to be found out.

[0246] Step 805: Perform matching analysis based on the solution information and the abnormal cause information stored in the preset solution database to determine the solution corresponding to the abnormal cause information, and define the solution as the current solution information.

[0247] The current solution information is information about the method to solve the problem corresponding to the abnormal cause information. The database stores the mapping relationship between the solution information and the abnormal cause information, which is matched and output by the staff in this field according to the existing situations in various logistics and the possible methods of normal processing. When the system receives the abnormal cause information, it automatically finds the corresponding solution from the database and outputs it as the current solution information.

[0248] Step 806: Output the abnormality cause information and the current solution information.

[0249] When customers come to ask what's going on, it's often a logistics problem. When outputting the abnormality reason information, customers not only need the reason, but also the solution to the reason, so the current solution information is also output. In this way, the customer's intention can be figured out without knowing the customer's true intention, thereby improving the intelligent service of customer service.

[0250] Reference Figure 9 , the method of calculating the actual difference information includes:

[0251] Step 900: Obtain historical vocabulary information corresponding to the current order number information.

[0252] The historical vocabulary information is the information of the questions reported by customers in the past corresponding to the same order number. This includes inquiries and other questions. The method of obtaining is to form a record every time a customer inquires.

[0253] Step 901: Determine whether there is preset return keyword information in the historical vocabulary information.

[0254] The return keyword information is the keyword information for returning the product to the merchant. The purpose of the judgment is to determine whether a return process is executed, thereby causing the start point information and the end point information to change.

[0255] Step 9011: If there is return keyword information, obtain the node that received the historical vocabulary information corresponding to the return keyword information, and define this node as the first new starting point information.

[0256] The first new starting point information is the information of the node that received the historical vocabulary information corresponding to the return keyword information. If there is return keyword information, it means that at the corresponding network point in the historical process, it turned back to the merchant corresponding to the starting point information. Then, the node that received the historical vocabulary information corresponding to the return keyword information is used as the first new starting point information.

[0257] Step 9012: If there is no return keyword information, determine whether there is the preset change keyword information in the historical vocabulary information.

[0258] The change keyword information is the information of the keyword for changing the address. If it does not exist, it means that the end point information does not change, and then re - judge. The purpose of the judgment is to determine whether there is a process of changing the address that causes the end point information to change.

[0259] Step 902: Perform matching analysis based on the node information, node time information, first new starting point information, and starting point information stored in the node database to determine the node and node time corresponding to the first new starting point information and the starting point information, define this node as the first new node information, and define this node time as the first new node time information.

[0260] The first new node information is the information of the node that exists when the starting point is updated to the first new starting point information and the end point is updated to the starting point information. The first new node time information is the information of the time that the first new node information theoretically stays. The database is automatically established by the logistics system. When the system receives the first new starting point information as the starting point and the starting point information as the end point, it automatically finds the corresponding node and node time from the database and outputs them as the first new node information and the first new node time information respectively.

[0261] Step 903: Update the first new node information to the pre - node information and recalculate the actual difference information.

[0262] The update method is that the information before the first new starting point information remains unchanged, that is, the nodes and node times between the starting point and the first new starting point information remain unchanged, while the nodes after the first new node information are changed to the first new node information, and then the actual difference information is recalculated.

[0263] Step 9041: If there is change keyword information, extract the first new end point information from the historical vocabulary information and obtain the node that received the historical vocabulary information corresponding to the change keyword information, and define this node as the second new starting point information.

[0264] The first new end point information is the information of the new end point input in the historical vocabulary information when the customer requests to change the keyword information. When the user needs to change the address, a new address must be entered. The second new start point information is the information of the node that receives the historical vocabulary information corresponding to the changed keyword information.

[0265] Step 9042: If there is no changed keyword information, directly calculate the actual difference information.

[0266] If there is no changed keyword information, it means that the start point information and the end point information remain unchanged at this time, so the actual difference information can be directly calculated. There may also be keywords such as delayed receipt of goods here, but it can also be directly calculated. At this time, only the abnormal reason information needs to be output.

[0267] Step 905: Perform a matching analysis based on the node information, node time information, second new start point information, and first new end point information stored in the node database to determine the node and node time corresponding to the second new start point information and the first new end point information. Define this node as the second new node information and define this node time as the second new node time information.

[0268] The second new node information is the information of the node that exists in the middle when the start point is updated to the second new start point information and the end point is updated to the first new end point information. The second new node time information is the information of the theoretically staying time of the second new node information. The database is automatically established by the logistics system. When the system receives the second new start point information as the start point and the first new end point information as the end point, it automatically finds the corresponding node and node time from the database and outputs them as the second new node information and the second new node time information respectively.

[0269] Step 906: Update the second new node information to the pre-node information and then recalculate the actual difference information.

[0270] The update method is that the information before the second new start point information remains unchanged, that is, the nodes and node times between the start point and the second new start point information remain unchanged, while the nodes after the second new node information are changed to the second new node information, and then the actual difference information is recalculated.

[0271] Based on the same inventive concept, an embodiment of the present invention provides an intelligent message processing system, including:

[0272] Refer to Figure 10 , an intelligent message processing system, including:

[0273] An information acquisition module 1003, configured to acquire merchant feedback information;

[0274] A processing module 1001, connected to the information acquisition module 1003 and the judgment module 1002, for storing and processing information;

[0275] An identification module 1004, connected to the processing module 1001, is used to identify keyword information in the merchant feedback information, where the keyword information includes the current order number information and the current intention information;

[0276] The processing module 1001 performs matching analysis based on the logistics status information stored in the preset logistics database and the current order number information to determine the logistics status corresponding to the current order number information, and defines this logistics status as the current logistics status information;

[0277] The processing module 1001 performs matching analysis based on the processing method information stored in the preset method database and the current intention information to determine the processing method corresponding to the current intention information, and defines this processing method as the current processing method information;

[0278] An output module 1005, connected to the processing module 1001, is used to output the current intention information;

[0279] A calculation module 1006, connected to the processing module 1001, is used to calculate the actual difference information;

[0280] An analysis module 1007, connected to the processing module 1001, is used to analyze the feedback object information in the current processing method information;

[0281] A judgment module 1002 is used to judge whether the feedback object information is the preset merchant information;

[0282] If the judgment module 1002 judges that it is, the processing module 1001 processes the feedback object information according to the current logistics status information using the current processing method information;

[0283] If the judgment module 1002 judges that it is not, the processing module 1001 performs matching analysis based on the logistics node information stored in the preset node database and the current logistics status information to determine the logistics node in the current logistics status information, and defines this logistics node as the current logistics node information;

[0284] The processing module 1001 sends the current processing method information to the network point corresponding to the current logistics node information and feedbacks the preset processed information to the merchant corresponding to the merchant feedback information.

[0285] The above are all preferred embodiments of the present application. Without limiting the protection scope of the present application accordingly, any feature disclosed in this specification (including the abstract and drawings), unless specifically described, can be replaced by other equivalent or similar-purpose alternative features. That is, unless specifically described, each feature is only an example of a series of equivalent or similar features.

Claims

1. An intelligent message processing method, characterized in that, Including: Obtain merchant feedback information; Identify keyword information in the merchant feedback information, where the keyword information includes the current order number information and the current intention information; Perform matching analysis based on the logistics status information stored in the preset logistics database and the current order number information to determine the logistics status corresponding to the current order number information, and define this logistics status as the current logistics status information; Perform matching analysis based on the processing method information stored in the preset method database and the current intention information to determine the processing method corresponding to the current intention information, and define this processing method as the current processing method information; Analyze the feedback object information in the current processing method information; Judge whether the feedback object information is the preset merchant information; If so, process the feedback object information according to the current logistics status information using the current processing method information; If not, perform matching analysis based on the logistics node information stored in the preset node database and the current logistics status information to determine the logistics node in the current logistics status information, and define this logistics node as the current logistics node information; Send the current processing method information to the network point corresponding to the current logistics node information and feedback the preset processed information to the merchant corresponding to the merchant feedback information.

2. The intelligent message processing method according to claim 1, wherein The method for identifying the current intention information in the keyword information of the merchant feedback information includes: Analyze the merchant feedback information to determine the vocabulary information in the merchant feedback information; Perform matching analysis based on the logistics keyword information stored in the preset keyword database and the vocabulary information to determine the logistics keyword stored in the vocabulary information, and define this logistics keyword as the effective logistics keyword information; Perform matching analysis based on the logistics vocabulary information stored in the preset vocabulary database and the effective logistics keyword information to determine the logistics vocabulary corresponding to the effective logistics keyword information, and define this logistics vocabulary as the effective logistics vocabulary information; Match the effective logistics vocabulary information with the vocabulary information one by one; If the match is successful, select the successfully matched effective logistics vocabulary information and define it as the first effective logistics vocabulary information; Perform matching analysis based on the intention information stored in the preset intention database and the first effective logistics vocabulary information to determine the intention corresponding to the first effective logistics vocabulary information, and define this intention as the first effective intention information; Output the first effective intention information as the current intention information; If the match is unsuccessful, output the preset package tracking operation information as the current intention information.

3. The intelligent message processing method according to claim 2, wherein The method for matching the effective logistics vocabulary information with the vocabulary information one by one includes: Screen out the character information in the effective logistics vocabulary information that is not the effective logistics keyword information, and define it as the effective character information; Remove the effective character information from the effective logistics vocabulary information to obtain new logistics vocabulary information; Judge whether the new logistics vocabulary information is the logistics vocabulary information stored in the vocabulary database; If so, output the new logistics vocabulary information and output the match success information; If not, screen out the character information in the vocabulary information that is not the effective logistics keyword information, and define it as the non-effective character information; Perform matching analysis based on the association degree information, valid character information, and invalid character information stored in the preset association database to determine the association degree of the valid character information and the invalid character information, and define this association degree as the valid association degree information; Judge whether the valid association degree information is greater than the preset association threshold information; If so, output the valid logistics vocabulary information and output the matching success information; If not, output the matching failure information.

4. An intelligent message processing method according to claim 2, characterized in that The method for outputting the current intention information includes: Obtain the negative term information in the merchant feedback information; Judge whether the negative term information exists; If it exists, judge whether the position information where the negative term information is located is before the first valid logistics vocabulary information; If so, perform matching analysis based on the intention information, the first valid logistics vocabulary information, and the negative term information stored in the intention database to determine the intention corresponding to the first valid logistics vocabulary information and the negative term information, and define this intention as the second valid intention information; Output the second valid intention information as the current intention information; If not, output the first valid intention information as the current intention information; If it does not exist, output the first valid intention information as the current intention information.

5. An intelligent message processing method according to claim 4, characterized in that The method for outputting the second valid intention information as the current intention information includes: Obtain the assumed term information in the merchant feedback information; Judge whether the assumed term information exists; If it exists, judge whether the position information where the assumed term information is located is before the negative term information; If so, update the second valid intention information to the first valid intention information and output it as the current intention information; If not, output the second valid intention information as the current intention information; If it does not exist, output the second valid intention information as the current intention information.

6. The intelligent message processing method according to claim 4, wherein The method for further outputting the current intention information includes: Obtain the quantity information of the current intention information in the merchant feedback information; Judge whether the quantity information is greater than 1; If it is equal to 1, output it with the current intention information; If it is greater than 1, obtain the valid vocabulary number information; Perform matching analysis based on the priority information and the valid vocabulary number information stored in the preset destination database to determine the priority corresponding to the valid vocabulary number information, and define this priority as the current priority information; Screen out the valid vocabulary number information with the largest current priority information and output the corresponding valid logistics vocabulary information.

7. The intelligent message processing method according to claim 2, wherein The method for processing the feedback object information according to the current logistics status information using the current processing method information includes: Perform matching analysis based on the possible process information and the current logistics status information stored in the preset process database to determine the possible process corresponding to the current logistics status information, and define this possible process as the current possible process information; Match the current possible process information with the current processing method information one by one; If the match is successful, output the current processing method information; If the match is unsuccessful, output the package tracking operation information as the current processing method information.

8. The intelligent message processing method according to claim 2, wherein If both the effective logistics vocabulary information and the vocabulary information do not match successfully, the method of outputting the package tracking operation information as the current processing method information includes: Match the preset query vocabulary information with the vocabulary information one by one; If the match is successful, analyze the starting point information and the ending point information in the current order number information; Perform matching analysis based on the node information and node time information stored in the preset node database, as well as the starting point information and the ending point information, to determine the nodes and node times corresponding to the starting point information and the ending point information. Define this node as the pre-node information and define this node time as the pre-node time information; Obtain the segmented node information and segmented node time information of the current logistics status information; Calculate the difference between the segmented node time information and the corresponding pre-node time information one by one, and define this difference as the actual difference information; Judge whether the actual difference information falls within the corresponding time threshold information; If it falls within, output the package tracking operation information as the current processing method information; If it does not fall within, filter out the segmented node information that does not fall within the time threshold information and obtain the abnormal cause information; Perform matching analysis based on the solution information stored in the preset solution database and the abnormal cause information to determine the solution corresponding to the abnormal cause information, and define this solution as the current solution information; Output the abnormal cause information and the current solution information; If the match is unsuccessful, output the package tracking operation information as the current processing method information.

9. An intelligent message processing method according to claim 8, characterized in that, The method of calculating the actual difference information includes: Obtain the historical vocabulary information corresponding to the current order number information; Judge whether there is the preset return keyword information in the historical vocabulary information; If there is return keyword information, obtain the node that receives the historical vocabulary information corresponding to the return keyword information, and define this node as the first new starting point information; Perform matching analysis based on the node information and node time information stored in the node database, as well as the first new starting point information and the starting point information, to determine the nodes and node times corresponding to the first new starting point information and the starting point information. Define this node as the first new node information and define this node time as the first new node time information; Update the first new node information to the pre-node information and recalculate the actual difference information; If there is no return keyword information, judge whether there is the preset change keyword information in the historical vocabulary information; If there is change keyword information, extract the first new ending point information from the historical vocabulary information and obtain the node that receives the historical vocabulary information corresponding to the change keyword information, and define this node as the second new starting point information; Perform matching analysis based on the node information and node time information stored in the node database, as well as the second new starting point information and the first new ending point information, to determine the nodes and node times corresponding to the second new starting point information and the first new ending point information. Define this node as the second new node information and define this node time as the second new node time information; Update the second new node information to the pre-node information and recalculate the actual difference information; If there is no change keyword information, directly calculate the actual difference information.

10. An intelligent message processing system, characterized in that, Include: An information acquisition module for acquiring merchant feedback information; A processing module connected to the information acquisition module and the judgment module for information storage and processing; An identification module connected to the processing module for identifying keyword information in the merchant feedback information, where the keyword information includes the current order number information and the current intention information; The processing module performs matching analysis based on the logistics status information stored in the preset logistics database and the current order number information to determine the logistics status corresponding to the current order number information, and defines this logistics status as the current logistics status information; The processing module performs matching analysis based on the processing method information stored in the preset method database and the current intention information to determine the processing method corresponding to the current intention information, and defines this processing method as the current processing method information; An analysis module connected to the processing module for analyzing the feedback object information in the current processing method information; A judgment module for judging whether the feedback object information is the preset merchant information; If the judgment module determines that it is, the processing module processes the feedback object information according to the current logistics status information using the current processing method information; If the judgment module determines that it is not, the processing module performs matching analysis based on the logistics node information stored in the preset node database and the current logistics status information to determine the logistics node in the current logistics status information, and defines this logistics node as the current logistics node information; The processing module sends the current processing method information to the network point corresponding to the current logistics node information and feedbacks the preset processed information to the merchant corresponding to the merchant feedback information.

Images