Text semantic understanding analysis method and system applied to smart campus platform

Abstract

The application relates to a text semantic understanding analysis method and system applied to a smart campus platform. In the application, a user-input campus interactive text is received, a historical interactive text set is called, semantic correlation processing is performed on the two to construct a context semantic chain, the associated campus resource state information is positioned based on the context semantic chain, and a resource matching result is generated, the structured semantic understanding result is generated according to the resource matching result and is pushed to the user. In addition, the context semantic chain and the query parameter are updated through the reception of the subsequent interactive text fed back by the user to optimize the semantic correlation processing. In this way, the semantic understanding ability and service accuracy of the smart campus platform to the user text are effectively improved.

Description

Technical Field

[0001] The present application relates to the technical field of text analysis and processing, and in particular to a text semantic understanding and analysis system applied to a smart campus platform. Background Art

[0002] In today's digital age, the construction and application of smart campus platforms are becoming increasingly widespread, bringing great convenience to all aspects of school teaching, management, and services. However, existing smart campus platforms have certain limitations when handling interactions between users and the system. Currently, most smart campus platforms can only perform simple keyword matching and regularized responses to text information entered by users. This processing method cannot deeply understand the semantics behind the user's text and the user's true demands, resulting in an inability to provide accurate and comprehensive services when faced with complex user needs. For example, when a user asks for detailed information about a course, including questions about the course's teaching progress, textbook usage, and related event arrangements, existing systems may only be able to match partial information based on keywords, without comprehensively considering the user's historical interactions and contextual semantics, and thus cannot accurately understand the user's complete demands.

[0003] Moreover, existing systems lack effective management and utilization of contextual semantics when processing user interactions. When a user interacts with the system multiple times, there may be semantic continuity and association between each interaction, but existing systems are unable to identify and utilize these relationships, making each interaction seem like an isolated event and unable to form a coherent interaction process. This not only affects the user experience, but also reduces the system's service efficiency and accuracy. In addition, when it comes to associating and matching campus resource status information, existing systems are often only able to perform simple text comparisons and are unable to accurately locate relevant campus resource status information based on the user's semantic demands. For example, when a user asks about the usage of a certain classroom, the system may simply search for records containing the classroom number, without considering the user's specific needs, such as whether they need to find usage within a specific time period.

[0004] To sum up, the existing smart campus platform has obvious deficiencies in text semantic understanding and interactive processing, making it difficult to better provide users with personalized and accurate services, and the utilization efficiency and management level of campus resources also need to be improved. Summary of the Invention

[0005] In view of the above, in order to at least partially address the deficiencies in the prior art, in a first aspect, an embodiment of the present application provides a text semantic understanding and analysis method applied to a smart campus platform, the method comprising:

[0006] Receive campus interaction text input by a user in the smart campus platform, and retrieve the user's historical interaction text set from the user interaction database of the smart campus platform based on the user identifier; the campus interaction text includes text content and the user identifier, and the historical interaction text set includes text content previously input by the user and corresponding interaction time information;

[0007] Performing semantic association processing on the text content of the campus interactive text and the historical interactive text set, identifying the semantic continuation relationship between the campus interactive text and each historical interactive text in the historical interactive text set, and constructing a contextual semantic chain based on the semantic continuation relationship;

[0008] Locating the associated campus resource status information from the campus resource status library of the smart campus platform based on the context semantic chain, and generating a resource matching result by comparing the text fragment in the context semantic chain with the description text of the campus resource status information;

[0009] Generate a structured semantic understanding result according to the resource matching result, and push the semantic understanding result to the user terminal corresponding to the user identifier;

[0010] Receive the subsequent interactive text fed back by the user terminal, add the subsequent interactive text to the context semantic chain and update the semantic dependency relationship between each fragment, and adjust the query parameters of the campus resource status library based on the updated context semantic chain.

[0011] In the second aspect, an embodiment of the present application also provides a text semantic understanding and analysis system applied to a smart campus platform, including a processor and a machine-readable storage medium, the machine-readable storage medium is connected to the processor, the machine-readable storage medium is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the machine-readable storage medium to implement the text semantic understanding and analysis method applied to the smart campus platform.

[0012] On the third aspect, an embodiment of the present application also provides a computer-readable storage medium, characterized in that it is used to store programs, instructions or codes, and when the programs, instructions or codes are executed by the processor, the text semantic understanding and analysis method applied to the smart campus platform is implemented.

[0013] In summary, the text semantic understanding analysis method and system applied to the smart campus platform provided by the embodiments of the present application can realize effective utilization of the historical interaction information of the user by receiving the campus interaction text input by the user and calling the historical interaction text set, and provide rich context information for subsequent semantic understanding. The semantic continuation relationship between the campus interaction text and the historical interaction text can be accurately identified by constructing the context semantic chain through semantic association processing, and a coherent semantic context is formed, so that the system can better understand the real demands of the user. The relevant campus resource state information is located based on the context semantic chain, and the resource matching result is generated, so as to ensure that the system can accurately find the relevant campus resource state according to the semantic demands of the user, and improve the accuracy and efficiency of resource matching. The structured semantic understanding result is generated and pushed to the user, and clear and comprehensive service information is provided for the user, for example, the user demand expression, the associated resource state description and the next step operation instruction, and the user experience is improved. The subsequent interaction text fed back by the user terminal is received, and the context semantic chain and the query parameter are updated, so as to continuously optimize the semantic association processing capability of the system, so that the smart campus platform continuously learns and improves in the continuous interaction with the user, and the adaptability and intelligence of the smart campus platform are improved. In this way, the text semantic understanding capability and the interactive service level of the smart campus platform are improved, and the rational use and efficient management of the campus resources are promoted.

[0014] Other features and advantages of the present application will be described in detail in the following specific embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creating laborious work on the basis of the above drawings.

[0016] In order to more completely understand the present application and its beneficial effects, the following will be described in conjunction with the drawings, wherein the same reference numerals in the following description represent the same parts.

[0017] Figure 1 It is the application scenario of the text semantic understanding analysis method provided by the embodiments of the present application.

[0018] Figure 2 It is a flowchart of the text semantic understanding analysis method applied to the smart campus platform provided by the embodiments of the present application.

[0019] Figure 3 It is a schematic diagram of the text semantic understanding analysis system applied to the smart campus platform provided by the embodiments of the present application. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the embodiments described are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0021] See also Figure 1 , Figure 1 : This is a schematic diagram of an application scenario of the text semantic understanding and analysis method applied to the smart campus platform provided in an embodiment of the present application. Among them, the application scenario includes a smart campus platform for information interaction and multiple user terminals. Among them, in this embodiment, the smart campus platform can be a server, server cluster, or other device with big data processing and storage capabilities, and the user terminal can be a portable user terminal computer or other device with data collection and transmission capabilities, which is not specifically limited in this embodiment. Multiple user terminals may include student terminals or teacher terminals used by students, etc., which is not limited in this embodiment.

[0022] like Figure 2 As shown, the method includes steps S110 to S150, which are described in detail below.

[0023] Step S110: Receive campus interaction text input by a user in the smart campus platform, and retrieve the user's historical interaction text set from the user interaction database of the smart campus platform based on the user ID. The campus interaction text includes text content and user ID, and the historical interaction text set includes the text content input by the user in the past and the corresponding interaction time information.

[0024] In this embodiment, S110 may include the following sub-steps S111-S115, which are described in detail below.

[0025] Step S111: Receive the campus interaction text sent by the user terminal through the text interaction interface of the smart campus platform, parse the data format of the campus interaction text, extract the text content and user identifier in the campus interaction text, and the user identifier is used to uniquely identify the user identity in the smart campus platform.

[0026] In the daily operation of the smart campus platform, users use the text interaction interface of the smart campus platform through their terminal devices, such as mobile phones and computers, to send campus interaction texts to the platform. The text interaction interface is an important channel for users to interact with the smart campus platform through text. It can receive text information in various formats. After receiving the campus interaction text sent by the user, as a possible example, the data format of the text can be parsed. For example, the campus interaction text sent by the user may be data in JSON format. As a possible example, it can be parsed according to the rules of JSON format to extract the text content and user ID. The user ID is a unique identifier. It can be the user's student number, faculty number, etc., which is used to accurately distinguish different users in the smart campus platform.

[0027] Step S112: Generate a query request based on the user identifier, wherein the query request includes the user identifier and a data range parameter, and the data range parameter is used to limit the time range for retrieving historical interaction texts.

[0028] After extracting the user ID, a query request can be generated based on it. This query request contains not only the user ID but also a data range parameter. The data range parameter limits the time range for retrieving historical interaction text. For example, the system can set the data range parameter to interaction records within the past month. In this way, when querying historical interaction text, only interaction records for the user within the past month will be retrieved. By setting the data range parameter, you can avoid retrieving excessive and useless historical data and improve query efficiency.

[0029] Step S113: Send the query request to the user interaction database of the smart campus platform, which stores historical interaction records of each user. Each historical interaction record includes user identification, interaction time information, input text content, and system feedback content.

[0030] After generating a query request, as a possible example, the query request can be sent to the user interaction database of the smart campus platform. This user interaction database is an important place to store historical user interaction records, which stores historical interaction information of each user. Each historical interaction record contains the user ID, interaction time information, input text content, and system feedback content. For example, a historical interaction record may show that a user entered a question about a course at a specific time, and the system provided corresponding feedback on the question.

[0031] Step S114: Receive the query results returned by the user interaction database, extract all historical interaction records matching the user identifier from the query results, filter out records that exceed the time range specified by the data range parameter, extract the input text content and interaction time information in the remaining records, and combine them to generate a historical interaction text set for the user.

[0032] After receiving the query request, the user interaction database will perform a query based on the user ID and data range parameters in the request and return the query results. After receiving the query results, all historical interaction records that match the user ID will first be extracted from the results. Then, the records will be filtered to exclude records that exceed the time range specified by the data range parameters. Finally, the input text content and interaction time information are extracted from the remaining records and combined together to generate the user's historical interaction text set. For example, if the data range parameter is limited to records within the last month, as a possible example, the records that exceed this time range in the query results can be filtered out, and only the records within the last month are retained, and the input text content and interaction time information of the records are combined into a historical interaction text set.

[0033] Step S115: sorting the historical interaction text set in ascending order according to the interaction time information to obtain a historical interaction text sequence arranged in chronological order, wherein each historical interaction text in the historical interaction text sequence contains corresponding interaction time information and input text content.

[0034] After generating a set of historical interaction texts, as a possible example, the text set can be sorted in ascending order according to the interaction time information. This results in a chronological sequence of historical interaction texts. In this sequence, each historical interaction text contains the corresponding interaction time information and input text content. The sorted historical interaction text sequence facilitates subsequent processing and analysis of historical interaction information and can better identify semantic continuity relationships between texts. For example, through the sorted sequence, the text content entered by users at different time points can be clearly seen, making it easier to analyze changes in users' interaction habits and semantic demands.

[0035] Step S120: Perform semantic association processing on the text content of the campus interactive text and the historical interactive text collection, identify semantic continuity relationships between the campus interactive text and each historical interactive text in the historical interactive text collection, and construct a contextual semantic chain based on the semantic continuity relationships. The contextual semantic chain includes related text segments sorted by interaction time and the semantic dependency relationships between each segment.

[0036] In this embodiment, S120 may include the following sub-steps S121-S125, which are described in detail below.

[0037] Step S121: performing word segmentation processing on the text content of the campus interactive text to obtain multiple current text word units, and performing word segmentation processing on the input text content of each historical interactive text in the historical interactive text set to obtain multiple historical text word units.

[0038] In order to perform semantic association processing, it is first necessary to perform word segmentation on the text content in the campus interactive text and historical interactive text sets. Word segmentation is to split the text content into independent word units. For campus interactive text, as a possible example, its text content can be segmented to obtain multiple current text word units. For example, if the campus interactive text is "I want to know the teaching progress of this course", after word segmentation, you may get "I", "want", "know", "this", "course", "teaching", "progress" and other current text word units. Similarly, for each historical interactive text in the historical interactive text set, the system will also perform word segmentation on its input text content to obtain multiple historical text word units.

[0039] Step S122: extracting the core noun unit and the core verb unit from the current text word unit, wherein the core noun unit represents the object requested by the user, and the core verb unit represents the behavior requested by the user.

[0040] After obtaining the current text word unit, as a possible example, core noun units and core verb units can be extracted from it. Core noun units represent the object of the user's request, and core verb units represent the behavior of the user's request. For example, in the current text word unit after the above word segmentation, "course" is a core noun unit, representing the object of the user's request; "understand" is a core verb unit, representing the user's request behavior. By extracting core noun units and core verb units, we can more accurately grasp the core of the user's request.

[0041] Step S123: traverse each historical interaction text in the historical interaction text set, extract the core noun unit and the core verb unit in the historical text word unit of the historical interaction text, and match them with the core noun unit and the core verb unit of the current text word unit.

[0042] As a possible example, each historical interactive text in the historical interactive text set can be traversed, and the historical text word units of each historical interactive text can be processed to extract the core noun units and core verb units in the historical interactive text. Then, the core noun units and core verb units are matched with the core noun units and core verb units of the current text word unit. For example, for a certain historical interactive text, after word segmentation and core unit extraction, the core noun unit "course" and the core verb unit "enroll" are obtained, which are compared with the core noun unit "course" and the core verb unit "understand" of the current text word unit.

[0043] Step S124: If the core noun unit of any historical interactive text is the same as the core noun unit of the current text word unit, and the core verb unit has a logical association, then it is determined that the campus interactive text and the historical interactive text have a semantic continuity relationship, and the logical association includes the succession relationship of actions, the progressive relationship of states, or the association relationship of objects.

[0044] In this embodiment, S124 may include the following sub-steps S1241-S1244, which are described in detail below.

[0045] Step S1241: performing a character string comparison on the core noun unit of the current text word unit and the core noun unit of any of the historical interactive texts. If the two are exactly the same, it is determined that the noun matching is successful.

[0046] When determining a semantic continuity relationship, the core noun unit of the current text word unit is first compared with the core noun unit of any of the historical interactive texts. If the two character strings are exactly the same, the noun match is determined to be successful. For example, if the core noun unit of the current text word unit is "课" and the core noun unit of the historical interactive text is also "课", then the noun match is successful.

[0047] Step S1242: If the noun is matched successfully, extract the core verb unit of the current text word unit and the core verb unit of the historical interactive text, and query the semantic distance between the two through the semantic association dictionary, which stores the semantic association relationship between verbs and the corresponding association strength.

[0048] When the noun matching is successful, as one possible example, the core verb unit of the current text word unit and the core verb unit of the historical interaction text can be extracted. Then, the semantic distance between the two core verb units is queried through a semantic association dictionary. The semantic association dictionary stores a large number of semantic association relationships between verbs and corresponding association strengths. For example, the semantic association dictionary can show that there is a certain semantic distance between “understand” and “enroll”, and this distance is determined according to the degree of their semantic association.

[0049] Step S1243: If the semantic distance is less than a preset threshold, it is determined that the core verb units have a logical association, and the type of the logical association is determined according to the association relationship in the semantic association dictionary; the action continuation relationship means that the action result of the previous verb is the action premise of the next verb, the state progression relationship means that the state described by the previous verb develops into the state described by the next verb, and the object association relationship means that the two verbs act on different aspects of the same object.

[0050] If the semantic distance obtained by the query is less than the preset threshold, it is determined that the core verb units have a logical association. The type of the logical association is determined according to the association relationship in the semantic association dictionary. For example, if the core verb unit of the current text word unit is “understand” and the core verb unit of the historical interaction text is “enroll”, the semantic association dictionary shows that there is an object association relationship between them, because the two verbs act on different aspects of the object “course”, one is to understand the course situation, and one is to enroll in the course.

[0051] Step S1244: If the noun matching is successful and the core verb units have a logical association, the noun matching result and the verb association result are integrated to determine that the campus interaction text and the historical interaction text have a semantic continuation relationship, and the interaction time information and the semantic dependency relationship type of the historical interaction text are recorded.

[0052] When the noun matching is successful and the core verb units have a logical association, as one possible example, the two results can be integrated to determine that the campus interaction text and the historical interaction text have a semantic continuation relationship. At the same time, the interaction time information and the semantic dependency relationship type of the historical interaction text are recorded. For example, if it is determined that the campus interaction text and a certain historical interaction text have a semantic continuation relationship, and the semantic dependency relationship type is an object association relationship, as one possible example, the interaction time of the historical interaction text and the semantic dependency relationship type of this object association can be recorded.

[0053] Step S125: Sort the historical interactive texts with semantic continuity relationships according to the interaction time information, and form a text sequence together with the campus interactive text. Analyze the semantic dependency between adjacent texts in the text sequence, and construct a context semantic chain based on the text sequence and the semantic dependency. The semantic dependency includes object dependency, conditional dependency, and result dependency. Each node in the context semantic chain corresponds to a text fragment, and the lines between nodes represent semantic dependencies.

[0054] In this embodiment, S125 may include the following sub-steps S1251-S1255, which are described in detail below.

[0055] Step S1251: Each text segment in the text sequence is a node, and the node attributes include text content, interaction time information and resource pointing nouns.

[0056] After obtaining the historical interaction texts with semantic continuity relationships, as a possible example, the historical interaction texts can be sorted according to the interaction time information and formed into a text sequence together with the campus interaction texts. Then, each text fragment in this text sequence is regarded as a node. Each node has corresponding attributes, including text content, interaction time information and resource pointing nouns. For example, a node may represent a historical interaction text, and its node attributes include the specific content of the text, the interaction time and the resource pointing nouns involved in the text, such as "course".

[0057] Step S1252: establishing directed edges between adjacent nodes according to the semantic dependency relationship, wherein the direction of the directed edges is from the node with an earlier time to the node with a later time, and the attributes of the directed edges include dependency type and dependency strength.

[0058] Based on the semantic dependencies between adjacent texts in a text sequence, as a possible example, directed edges can be established between adjacent nodes. Directed edges point from earlier nodes to later nodes, reflecting the temporal order of text interactions. Directed edges have properties, including dependency type and dependency strength. For example, if an object dependency exists between two adjacent nodes, as a possible example, a directed edge can be established between the two nodes, with the dependency type marked as object dependency and the dependency strength determined based on the association strength in the semantic association dictionary.

[0059] Step S1253: normalizing the dependency strengths of the directed edges so that the dependency strengths of all directed edges are within a normalized range, wherein the initial value of the dependency strength is determined according to the association strength in the semantic association dictionary.

[0060] To facilitate subsequent processing and analysis, as a possible example, the dependency strengths of directed edges can be normalized. The initial value of the dependency strength is determined based on the association strength in the semantic association dictionary. Normalization can bring the dependency strengths of all directed edges into a uniform range, for example, normalizing the dependency strengths to between 0 and 1. This prevents large differences in the values ​​of different dependency strengths, which could affect subsequent calculations and analysis.

[0061] Step S1254: Detect whether there are non-adjacent nodes in the text sequence but have a semantic dependency relationship. If so, add a cross-node directed edge between the nodes, and the dependency strength of the cross-node directed edge is lower than the dependency strength of the adjacent node directed edges.

[0062] As a possible example, it is possible to detect whether there are non-adjacent nodes in a text sequence that have a semantic dependency relationship. If this is the case, as a possible example, a cross-node directed edge can be added between the nodes. Since the semantic dependency between non-adjacent nodes is relatively weak, the dependency strength of the cross-node directed edge is lower than the dependency strength of the adjacent node directed edge. For example, if there is a semantic association between two non-adjacent nodes, but this association is not as close as the association between adjacent nodes, as a possible example, a cross-node directed edge can be added and a lower dependency strength can be set.

[0063] Step S1255: All nodes and directed edges are combined to form a contextual semantic chain including node attributes and edge attributes. The contextual semantic chain stores the semantic association relationship between each text segment in a graphical structure.

[0064] Finally, as a possible example, all nodes and directed edges can be combined to form a contextual semantic chain. This contextual semantic chain contains node and edge attributes, storing the semantic relationships between text segments in a graphical structure. Through this contextual semantic chain, the system can more intuitively understand the semantic relationships between texts, providing strong support for subsequent resource location and semantic understanding.

[0065] Step S130: Based on the context semantic chain, the related campus resource status information is located from the campus resource status library of the smart campus platform. The campus resource status library contains the teaching resource usage status, facility operation status, activity development status and transaction progress information. By comparing the text fragments in the context semantic chain with the descriptive text of the campus resource status information, a resource matching result is generated.

[0066] In this embodiment, S130 may include the following sub-steps S131-S135, which are described in detail below.

[0067] Step S131: Parse the contextual semantic chain and extract resource-directed nouns from each text segment. The resource-directed nouns represent the campus resources involved in the user's request. For example, resource-directed nouns related to teaching resources include course name, classroom number, and textbook name; resource-directed nouns related to facilities include dormitory number, equipment name, and venue name; resource-directed nouns related to activities include activity name, club name, and event name; and resource-directed nouns related to affairs include matter name, certificate type, and application number.

[0068] As a possible example, the context semantic chain can be parsed to extract resource-pointing nouns from each text fragment. Resource-pointing nouns represent the campus resources involved in the user's demands. Different types of campus resources correspond to different resource-pointing nouns. For example, in terms of teaching resources, the resource-pointing noun may be the course name, classroom number, or textbook name; in terms of facilities, it may be the dormitory number, equipment name, or venue name; in terms of activities, it may be the activity name, club name, or event name; in terms of affairs, it may be the matter name, certificate type, or application number. By extracting resource-pointing nouns, the specific campus resources involved in the user's demands can be clarified.

[0069] Step S132: Determine the target resource type based on the resource-pointing noun. For example, a teaching resource-pointing noun corresponds to the teaching resource usage status, a facility-related resource-pointing noun corresponds to the facility operation status, an activity-related resource-pointing noun corresponds to the activity development status, and a transaction-related resource-pointing noun corresponds to the transaction processing progress information.

[0070] As one possible example, the target resource type can be determined based on the extracted resource-pointing noun. Different resource-pointing nouns correspond to different target resource types. For example, if the resource-pointing noun is a course name, the target resource type is the teaching resource usage status; if it is a dormitory number, the target resource type is the facility operation status. This allows the user to accurately locate the type of campus resource status information required by the user.

[0071] Step S133: Access the corresponding sub-library of the campus resource status library of the smart campus platform according to the target resource type.

[0072] Among them, the teaching resource usage status sub-library stores the current teaching status of the course, classroom occupancy, and textbook inventory information; the facility operation status sub-library stores the maintenance status of the dormitory, equipment operating parameters, and venue opening hours; the activity status sub-library stores the registration progress of the activity, the activity schedule of the club, and the process of holding the competition; the transaction progress information sub-library stores the approval stage of the matter, the status of the certificate processing, and the processing progress of the application.

[0073] After determining the target resource type, as an example, you can access the corresponding sub-library of the Smart Campus Platform's Campus Resource Status Library based on that type. The Campus Resource Status Library contains multiple sub-libraries, each storing different types of campus resource status information. For example, if the target resource type is teaching resource usage status, as an example, you can access the Teaching Resource Usage Status sub-library, which stores information such as the current course teaching status, classroom occupancy, and textbook inventory.

[0074] Step S134: searching the corresponding sub-library for resource status records containing the resource pointing noun, extracting the description text of each resource status record, and performing a semantic similarity comparison between the description text and the input text content of the corresponding text segment in the context semantic chain.

[0075] In this embodiment, S134 may include the following sub-steps S1341-S1346, which are described in detail below.

[0076] Step S1341: performing word segmentation processing on the description text and the input text content of the corresponding text segment in the context semantic chain respectively to obtain a description word unit sequence and a text word unit sequence.

[0077] In order to compare semantic similarity, we first need to perform word segmentation on the input text content of the description text and the corresponding text fragment in the context semantic chain. Through word segmentation, the description text is split into a sequence of description word units, and the input text content is split into a sequence of text word units. For example, if the description text is "The course is currently being taught", after word segmentation, the description word unit sequence "the" "course" "currently" "being" "taught" "in" is obtained; if the input text content of the corresponding text fragment is "I want to know about the teaching situation of this course", after word segmentation, the text word unit sequence "I" "want" "to know" "the" "teaching" "situation" of "this" "course" is obtained.

[0078] Step S1342: extract key descriptive words from the descriptive word unit sequence, wherein the key descriptive words are adjective units and verb units that describe resource status; extract key appeal words from the text word unit sequence, wherein the key appeal words are adjective units and verb units that express user needs.

[0079] After obtaining the descriptive word unit sequence and the text word unit sequence, as a possible example, key descriptive words and key appeal words can be extracted from them. Key descriptive words are adjective units and verb units that describe resource status, while key appeal words are adjective units and verb units that express user needs. For example, in the above descriptive word unit sequence, "teaching" is a key descriptive word; in the text word unit sequence, "understanding" and "teaching" are key appeal words.

[0080] Step S1343: Calculate the overlap between the key description words and the key appeal words, where the overlap is the ratio of the number of identical word units to the total number of word units.

[0081] As a possible example, the degree of overlap between key descriptors and key appeal words can be calculated. The overlap is calculated as the ratio of the number of identical word units to the total number of word units. For example, if the key descriptor is "teaching," and the key appeal words are "understanding" and "teaching," the identical word unit is "teaching," and the total number of word units is four, including two key descriptors and two key appeal words, then the overlap is 1 / 4.

[0082] Step S1344: converting the description text and the input text content into semantic vectors through a semantic vector model, and calculating the cosine similarity of the two semantic vectors.

[0083] As one possible example, a semantic vector model can be used to convert the description text and input text content into semantic vectors. The semantic vector model converts text information into vector representations for similarity calculation. The cosine similarity of the two semantic vectors is then calculated. Cosine similarity measures the cosine value of the angle between two vectors. The closer the cosine value is to 1, the more similar the two vectors are, indicating that the semantics of the description text and input text content are more similar.

[0084] Step S1345: Perform weighted summation on the overlap and the cosine similarity to obtain comprehensive semantic similarity. The weight of the overlap and the weight of the cosine similarity are dynamically adjusted according to the resource type. The weight of the overlap in the teaching resource type is higher than the weight of the cosine similarity, and the weight of the cosine similarity in the facility resource type is higher than the weight of the overlap.

[0085] As a possible example, the weighted sum of overlap and cosine similarity can be used to obtain comprehensive semantic similarity. The weights of overlap and cosine similarity are dynamically adjusted based on the resource type. For teaching resource types, the weight of overlap is higher than that of cosine similarity because keyword matching is more important in teaching resource descriptions. For facility resource types, the weight of cosine similarity is higher than that of overlap because facility resource descriptions may focus more on expressing overall semantics.

[0086] Step S1346: taking the comprehensive semantic similarity as the semantic similarity comparison result between the description text and the input text content.

[0087] Finally, the calculated comprehensive semantic similarity is taken as the semantic similarity comparison result of the description text and the input text content. This result can reflect the semantic similarity between the description text and the input text content, and provide a basis for subsequent resource matching.

[0088] Step S135: The resource state record with the highest semantic similarity is determined as the associated campus resource state information. The state description field, the responsible person information field, and the update time field in the campus resource state information are extracted, combined to generate a resource matching result. If no resource state record containing the resource pointing noun is retrieved, a resource non-matching prompt information is generated.

[0089] According to the semantic similarity comparison result, as a possible example, the resource state record with the highest semantic similarity can be determined as the associated campus resource state information. Then, the state description field, the responsible person information field, and the update time field in the campus resource state information are extracted, and the fields are combined to generate a resource matching result. For example, if the associated campus resource state information is about a course, the state description field can be the current teaching progress of the course, the responsible person information field can be the contact information of the teaching teacher, and the update time field can be the last update time of the information. If no resource state record containing the resource pointing noun is retrieved in the corresponding sub-library, as a possible example, a resource non-matching prompt information can be generated to inform the user that no relevant campus resource state information is found.

[0090] Step S140: A structured semantic understanding result is generated according to the resource matching result, the semantic understanding result containing a user appeal expression, an associated resource state description, and a next step operation guide. The semantic understanding result is pushed to the user terminal corresponding to the user identifier.

[0091] In this embodiment, S140 can include the following sub-steps S141-S146, which are described in detail below.

[0092] Step S141: The resource matching result is analyzed, and the state description field, the responsible person information field, and the update time field in the associated campus resource state information are extracted. If the resource matching result is a resource non-matching prompt information, the state description field is a resource non-existent description, the responsible person information field is empty, and the update time field is the current time.

[0093] As a possible example, the resource matching results can be parsed to extract the status description field, responsible person information field, and update time field from the associated campus resource status information. If the resource matching result indicates that the resource is not matched, the status description field will display an explanation that the resource does not exist, the responsible person information field will be empty, and the update time field will be the current time. For example, if the resource matching result indicates that no relevant information for a certain course was found, the status description field will display "Course Information Not Found", the responsible person information field will be empty, and the update time field will be the current system time.

[0094] Step S142: Based on the text content of the campus interactive text in the contextual semantic chain, the core demands of the user are extracted and a user demand expression is generated. The user demand expression is a standardized description of the text content input by the user, including the demand object and the demand behavior.

[0095] As an example, based on the textual content of campus interaction text within the contextual semantic chain, we can extract the user's core demands and generate a user demand statement. A user demand statement is a standardized description of the user's input text content, including the demand object and demand behavior. For example, if the campus interaction text is "I want to learn about the teaching progress of this course," the user demand statement might be "The user wants to learn about the teaching progress of this course," where "course" is the demand object and "learn about the teaching progress" is the demand behavior.

[0096] Step S143: Arrange the content of the status description field into a description of the associated resource status, wherein the description of the associated resource status includes the current status of the resource, the last update time, and the contact information of the relevant person in charge.

[0097] As a possible example, the contents of the Status Description field can be organized to form a Related Resource Status Description. This related resource status description includes the resource's current status, the time of its last update, and the contact information of the responsible person. For example, if the Status Description field indicates that a course is currently halfway through and was last updated yesterday, and the Responsible Person Information field displays the instructor's contact number, the Related Resource Status Description would display "This course is currently halfway through, was last updated yesterday, and the instructor's contact number is [XXXXXXXX]."

[0098] Step S144: Generate next steps based on the resource status description and user request. For example, if the resource status is normal and meets the user's request, the next steps may be resource acquisition or processing. If the resource status is abnormal or does not meet the user's request, the next steps may be resource recommendations or problem feedback.

[0099] As a possible example, the next step of the operation guide can be generated based on the description of the associated resource status and the user's request. If the resource status is normal and meets the user's needs, the operation guide may be the resource acquisition method or the processing entrance. For example, if the user wants to know the teaching progress of a certain course, the description of the associated resource status shows that the course teaching progress is normal and relevant information has been provided, then the operation guide may be "You can obtain detailed teaching materials for this course through [specific channel]." If the resource status is abnormal or does not meet the user's needs, the operation guide may be an alternative resource recommendation or a problem feedback channel. For example, if the course teaching progress is delayed, the operation guide may be "The teaching progress of this course is delayed, you can consider choosing [alternative course], or you can feedback the problem through [feedback channel]."

[0100] Step S145: Combine the user demand expression, the description of the associated resource status and the next operation instructions according to a preset structure to generate a structured semantic understanding result. The preset structure is that the user demand expression is in the front, the description of the associated resource status is in the middle, and the next operation instructions are in the back.

[0101] As a possible example, the user's request statement, the description of the associated resource status, and the next step operation instructions can be combined according to a preset structure to generate a structured semantic understanding result. The preset structure is that the user's request statement comes first, the description of the associated resource status is in the middle, and the next step operation instructions come last. For example, the structured semantic understanding result may be "The user wants to know the teaching progress of this course. The course is currently halfway through, the last update time was yesterday, and the instructor's contact number is [specific phone number]. You can obtain detailed teaching materials for this course through [XXX channel]."

[0102] Step S146: query the user terminal association table of the smart campus platform based on the user identifier, obtain the user terminal address corresponding to the user identifier, and push the semantic understanding result to the user terminal corresponding to the user terminal address through the platform message push service.

[0103] As one possible example, the user terminal association table of the smart campus platform can be queried based on the user ID to obtain the user terminal address corresponding to the user ID. The structured semantic understanding results can then be pushed to the user terminal corresponding to the user terminal address via the platform's message push service. For example, if the user terminal address corresponding to the user ID is a mobile phone number, as one possible example, the semantic understanding results can be sent to the mobile phone corresponding to the mobile phone number via SMS push service.

[0104] Step S150: Receive the subsequent interactive text fed back by the user terminal, add the subsequent interactive text to the context semantic chain and update the semantic dependency between each fragment, and adjust the query parameters of the campus resource status library based on the updated context semantic chain to optimize the accuracy of the next semantic association processing.

[0105] In this embodiment, S150 may include the following sub-steps S151-S155, which are described in detail below.

[0106] Step S151: monitoring the subsequent interactive text sent by the user terminal corresponding to the user identifier within a preset feedback time window. The subsequent interactive text is the user's response to the semantic understanding result, including confirmation expression, supplementary expression or correction expression.

[0107] As a possible example, the subsequent interaction text sent by the user terminal corresponding to the user identifier can be monitored within a preset feedback time window. The subsequent interaction text is the user's response to the semantic understanding result and may include confirmation statements, supplementary statements, or correction statements. For example, the user may send a confirmation statement such as "Confirm, I understand," or a supplementary statement such as "I would also like to know the exam schedule for this course," or a correction statement such as "My previous question was incorrect, I would like to know about another course."

[0108] Step S152: Perform word segmentation processing on the subsequent interactive text to obtain subsequent text word units, extract core noun units and core verb units in the subsequent text word units, and determine the semantic dependency relationship between the subsequent interactive text and the last text segment in the context semantic chain based on the core noun units and core verb units.

[0109] As a possible example, the subsequent interactive text can be segmented to obtain subsequent text word units. Then, core noun units and core verb units are extracted from them. Based on the core units, the semantic dependency relationship between the subsequent interactive text and the last text segment in the context semantic chain is determined. For example, if the subsequent interactive text is "I also want to know the exam schedule for this course", after segmentation, the core noun unit "course" and the core verb unit "understand" are obtained. By comparing with the last text segment in the context semantic chain, it is determined that there is a complementary semantic dependency relationship between them.

[0110] Step S153: The subsequent interaction text is added as a new node to the end of the contextual semantic chain. The interaction time information of the node is recorded as the time when the subsequent interaction text was received. A semantic dependency relationship is established between the new node and the previous node. The semantic dependency relationship is determined based on the type of the subsequent interaction text. The corresponding result dependency is confirmed, the corresponding condition dependency is supplemented, and the corresponding object dependency is modified.

[0111] As a possible example, the subsequent interaction text can be added to the end of the context semantic chain as a new node. The interaction time information of the node is recorded as the time of receiving the subsequent interaction text. According to the type of the subsequent interaction text, the semantic dependency relationship between the new node and the previous node is established. For example, if the subsequent interaction text is a confirmation expression, the semantic dependency relationship between the new node and the previous node is a result dependency; if it is a supplementary expression, it is a conditional dependency; if it is a correction expression, it is an object dependency.

[0112] Step S154: traversing the updated context semantic chain, recalculating the strength of the semantic dependency relationship between nodes, which is dynamically adjusted according to the interaction time interval and semantic similarity. The shorter the time interval and the higher the semantic similarity, the greater the dependency relationship strength.

[0113] As a possible example, the strength of the semantic dependency relationship between nodes can be recalculated by traversing the updated context semantic chain. The dependency relationship strength is dynamically adjusted according to the interaction time interval and semantic similarity. The shorter the time interval and the higher the semantic similarity, the greater the dependency relationship strength. For example, if the interaction time interval of two nodes is short and their text content is very similar in semantics, the dependency relationship strength between them will be greater.

[0114] Step S155: based on the resource pointing nouns in the updated context semantic chain and the strength of the semantic dependency relationship, adjusting the query parameters of the campus resource state library, increasing the retrieval weight of the resource pointing nouns corresponding to the high dependency relationship strength, and expanding the retrieval range of the related resource state records, to optimize the accuracy of resource positioning in the next semantic association processing.

[0115] In this embodiment, S155 can include the following sub-steps S1551-S1556, which will be described in detail below.

[0116] Step S1551: extracting the resource pointing nouns of all nodes from the updated context semantic chain, and counting the frequency of occurrence of each resource pointing noun and the sum of the corresponding node dependency strength.

[0117] As a possible example, the resource pointing nouns of all nodes can be extracted from the updated context semantic chain. Then, the frequency of occurrence of each resource pointing noun and the sum of the corresponding node dependency strength are counted. For example, if in the updated context semantic chain, the resource pointing noun "course" appears 3 times, and the sum of the node dependency strength related to it is a certain value, as a possible example, the information can be recorded.

[0118] Step S1552: multiply the frequency by the sum of the node dependency strengths to obtain a comprehensive weight of each resource-pointing noun, where the comprehensive weight represents the importance of the resource-pointing noun in the context semantic chain.

[0119] As a possible example, the frequency of each resource-pointing noun can be multiplied by the sum of the corresponding node dependency strengths to obtain a comprehensive weight. The comprehensive weight indicates the importance of the resource-pointing noun in the contextual semantic chain. For example, if the frequency of "course" is 3 and the sum of the node dependency strengths is a specific value, multiplying them together will give the comprehensive weight of "course".

[0120] Step S1553: Sort the resource-pointing nouns according to the comprehensive weights, select the first several resource-pointing nouns with the highest comprehensive weights as core search terms, and the rest as auxiliary search terms.

[0121] As a possible example, resource-directed nouns can be sorted based on their comprehensive weights. The top few resource-directed nouns with the highest comprehensive weights are selected as core search terms, and the remaining ones are used as auxiliary search terms. For example, if there are resource-directed nouns such as "course," "classroom," and "textbook," after sorting them based on their comprehensive weights, "course" and "classroom," which have the highest comprehensive weights, are selected as core search terms, and "textbook" is used as an auxiliary search term.

[0122] Step S1554: Adjust the query parameters of the campus resource status library, set the core search term as an exact matching condition, set the auxiliary search term as a fuzzy matching condition, and the matching weight of the core search term is higher than the matching weight of the auxiliary search term.

[0123] As a possible example, you can adjust the query parameters of the campus resource status database, setting the core search term as an exact match condition and the auxiliary search term as a fuzzy match condition. The matching weight of the core search term will be higher than the matching weight of the auxiliary search term. For example, when querying the campus resource status database, for the core search term "course", an exact match will be performed, and only records that completely match the name of "course" will be considered; for the auxiliary search term "textbook", a fuzzy match will be performed, and any record containing information related to "textbook" will be considered. The core search term has a higher matching weight and will be given priority in the matching process.

[0124] Step S1555: Expand the search time range of the resource status records corresponding to the core search term to include the most recently updated resource status records and historical related records to improve the comprehensiveness of resource positioning. The search time range of the auxiliary search term is limited to the most recently updated resource status records.

[0125] To improve the comprehensiveness of resource locating, as a possible example, the search time range for resource status records corresponding to the core search term can be expanded to include recently updated resource status records and historical related records. The search time range for auxiliary search terms is limited to recently updated resource status records. For example, the core search term "course" will search for the most recent teaching status records and historical teaching records for that course; the auxiliary search term "textbook" will only search for the most recent textbook inventory information.

[0126] Step S1556: The adjusted query parameters are stored in the query parameter configuration table for use in the next semantic association processing of the campus interactive text based on the user identifier.

[0127] Finally, as a possible example, the adjusted query parameters can be stored in a query parameter configuration table, and the parameters will be used for the next semantic association processing of campus interactive text based on the user ID. This can continuously optimize the query process and improve the accuracy of semantic association processing.

[0128] It should be understood that in the entire process of the above embodiment, if the user interaction text, historical interaction records and other data involved involve user privacy-sensitive data. In order to protect the privacy of the data and prevent leakage, this embodiment will adopt the following technical means to protect the security of the data. For example, first, the user data is encrypted and stored, and the data is encrypted using advanced encryption algorithms. Only authorized system modules can decrypt and access it. Secondly, during the data transmission process, a secure transmission protocol, such as the SSL / TLS protocol, is adopted to ensure that the data is not stolen or tampered with during the transmission process. In addition, strict permission management is implemented for access to the system. Only authorized personnel can access and process user data, and detailed log records are kept for access operations for auditing and tracking. At the same time, the system is regularly scanned and repaired for security vulnerabilities to promptly discover and deal with potential security risks and ensure the security of user privacy data.

[0129] In addition, in the application of the smart campus platform, if relevant data collection is involved in this embodiment, relevant regulations will be strictly followed and necessary technical measures will be taken to ensure compliance.

[0130] For example, before data collection begins, the system clearly informs users of the purpose, scope, method, and duration of data collection, obtaining their explicit authorization. By prominently displaying a detailed data collection statement on the Smart Campus platform, users are ensured to fully understand and independently decide whether to consent to providing data. Furthermore, a convenient authorization interface is provided, making it easy for users to authorize and cancel data.

[0131] Secondly, anonymization and de-identification technologies are employed during data collection. Sensitive user information, such as ID numbers and contact information, is encrypted upon collection and converted into anonymous identifiers. Subsequent data processing and analysis utilize only these anonymous identifiers, avoiding direct access to sensitive user information and thus reducing the risk of privacy breaches.

[0132] Furthermore, a strict data access control mechanism should be established. Only authorized personnel and system modules can access and process collected data. By setting different access rights levels, different personnel can only access the data required for their work, preventing data misuse and leakage. Furthermore, detailed logging of data access operations should be maintained for audit and traceability.

[0133] In addition, we regularly conduct security assessments and compliance checks on our data collection systems. We utilize third-party safety and regulatory agencies to conduct vulnerability scans and risk assessments on the platform system, promptly identifying and remediating potential security risks. We also closely monitor updates and changes to relevant laws and regulations, and promptly adjust our data collection strategies and technical methods to ensure compliance with the latest compliance requirements. Through the comprehensive application of these technical methods, the necessary data collection process not only meets the requirements of the Smart Campus Platform's text semantic understanding and analysis, but also fully protects the legitimate rights and interests of users, ensuring that data collection activities comply with relevant laws and regulations.

[0134] On the basis of the above, if Figure 3 The figure shows a schematic diagram of a text semantic understanding and analysis system applied to a smart campus platform provided by an embodiment of the present application. The system includes components such as a processor, a machine-readable storage medium, an input and output device, the machine-readable storage medium is connected to the processor, the machine-readable storage medium is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the machine-readable storage medium to implement the above-mentioned text semantic understanding and analysis method applied to the smart campus platform. Among them, the text semantic understanding and analysis system applied to the smart campus platform can be understood as the present application Figure 1 A part of the user terminal in the application scenario shown, or the user terminal itself.

[0135] The machine readable storage medium can be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read-Only memory (PROM), an Erasable Programmable Read-Only Memory (EPROM), an Electric Erasable Programmable Read-Only Memory (EEPROM) and the like. The machine readable storage medium is used to store a program, and the processor executes the program after receiving an execution instruction.

[0136] The processor can be an integrated circuit chip having a processing capability of signals. The processor can be, but is not limited to, a general processor, including a Central Processing Unit (CPU), a Network Processor (NP), a Digital Signal Processor (DSP), an Application-Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA) and the like.

[0137] To sum up, the text semantic understanding analysis method and system applied to the smart campus platform provided by the embodiments of the present application can effectively utilize the historical interaction information of the user by receiving the campus interactive text input by the user and calling the historical interactive text set, and provide rich context information for subsequent semantic understanding. The context semantic chain is constructed through semantic association processing, which can accurately identify the semantic continuation relationship between the campus interactive text and the historical interactive text, form a coherent semantic context, and enable the system to better understand the real demands of the user. Based on the context semantic chain, the associated campus resource state information is located and the resource matching result is generated, which ensures that the system can accurately find the relevant campus resource state according to the semantic demands of the user, and improves the accuracy and efficiency of resource matching. The structured semantic understanding result is generated and pushed to the user, which provides clear and comprehensive service information to the user, for example, can include user demand expression, associated resource state description and next step operation guidance, and improves the user experience. The subsequent interactive text fed back by the user terminal is received and the context semantic chain and the query parameter are updated, which can continuously optimize the semantic association processing capability of the system, so that the smart campus platform continuously learns and improves in the continuous interaction with the user, and improves the adaptability and intelligence of the smart campus platform. In this way, the text semantic understanding capability and the interactive service level of the smart campus platform are improved, and the rational use and efficient management of the campus resources are promoted.

[0138] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. The embodiments, implementation methods and related technical features of this application can be combined and replaced with each other in the absence of conflict. The above are only preferred embodiments of this application and are not intended to limit this application in any form. However, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of this application without departing from the content of the technical solution of this application are still within the scope of the technical solution of this application.

Claims

1. A text semantic understanding and analysis method applied to a smart campus platform, characterized in that: The method comprises: Receive campus interaction text input by a user in the smart campus platform, and retrieve the user's historical interaction text set from the user interaction database of the smart campus platform based on the user identifier; the campus interaction text includes text content and the user identifier, and the historical interaction text set includes text content previously input by the user and corresponding interaction time information; The text content of the campus interactive text is semantically associated with the historical interactive text set, the semantic continuation relationship between the campus interactive text and each historical interactive text in the historical interactive text set is identified, and a context semantic chain is constructed based on the semantic continuation relationship, wherein the context semantic chain is constructed in the following manner: each text segment in the text sequence is taken as a node, and the node attributes include text content, interaction time information and resource pointing nouns, and the text sequence is composed of the historical interactive text and the campus interactive text; directed edges are established between adjacent nodes according to the semantic dependency relationship between adjacent texts in the text sequence, and the direction of the directed edge is from the node with an earlier time to the node with an earlier time. For nodes with later time, the attributes of the directed edges include the dependency type and dependency strength; the dependency strength of the directed edges is normalized so that the dependency strength of all directed edges is within the normalized range, and the initial value of the dependency strength is determined according to the association strength in the semantic association dictionary; detecting whether there are non-adjacent nodes in the text sequence but with semantic dependency relationships, and if so, adding cross-node directed edges between the nodes, wherein the dependency strength of the cross-node directed edges is lower than the dependency strength of the adjacent node directed edges; combining all nodes and directed edges to form a contextual semantic chain including node attributes and edge attributes, wherein the contextual semantic chain stores the semantic association relationship between each text segment in a graphical structure; Locating the associated campus resource status information from the campus resource status library of the smart campus platform based on the context semantic chain, and generating a resource matching result by comparing the text fragment in the context semantic chain with the description text of the campus resource status information; Generate a structured semantic understanding result based on the resource matching result, and push the semantic understanding result to the user terminal corresponding to the user identifier, specifically including: parsing the resource matching result, extracting the status description field, the person in charge information field and the update time field in the associated campus resource status information; if the resource matching result is a resource unmatched prompt message, the status description field is a description of the resource not existing, the person in charge information field is empty, and the update time field is the current time; based on the text content of the campus interactive text in the context semantic chain, extract the core demands of the user and generate a user demand expression, which is a standardized description of the user input text content, including the demand object and demand behavior; organize the content of the status description field into an associated resource status description , the associated resource status description includes the current status of the resource, the most recent update time and the contact information of the relevant person in charge; based on the associated resource status description and the user's demand expression, the next operation guide is generated. If the resource status is normal and meets the user's demand, the operation guide is the resource acquisition path or processing entrance; if the resource status is abnormal or does not meet the user's demand, the operation guide is an alternative resource recommendation or a problem feedback path; the user demand expression, the associated resource status description and the next operation guide are combined according to a preset structure to generate a structured semantic understanding result; based on the user identifier, the user terminal association table of the smart campus platform is queried to obtain the user terminal address corresponding to the user identifier, and the semantic understanding result is pushed to the user terminal corresponding to the user terminal address through the platform message push service; Receive the subsequent interactive text fed back by the user terminal, add the subsequent interactive text to the context semantic chain and update the semantic dependency relationship between each fragment, and adjust the query parameters of the campus resource status library based on the updated context semantic chain.

2. The text semantic understanding and analysis method applied to the smart campus platform according to claim 1 is characterized in that: The receiving of campus interaction text input by a user in the smart campus platform and retrieving a historical interaction text set of the user from a user interaction database of the smart campus platform based on the user identifier includes: Receiving the campus interactive text sent by the user terminal through the text interactive interface of the smart campus platform, parsing the data format of the campus interactive text, and extracting the text content and user identifier in the campus interactive text; Generate a query request based on the user identifier, the query request including the user identifier and a data range parameter, the data range parameter being used to limit a time range for retrieving historical interaction texts; Sending the query request to a user interaction database of the smart campus platform, wherein the user interaction database stores historical interaction records of each user; Receive a query result returned by the user interaction database, extract all historical interaction records matching the user identifier from the query result, filter out records that exceed the time range defined by the data range parameter, extract input text content and interaction time information from the remaining records, and combine them to generate a historical interaction text set for the user; The historical interaction text set is sorted in ascending order according to the interaction time information to obtain a historical interaction text sequence arranged in chronological order, wherein each historical interaction text in the historical interaction text sequence includes corresponding interaction time information and input text content.

3. The text semantic understanding and analysis method applied to the smart campus platform according to claim 1 is characterized in that: The step of performing semantic association processing on the text content of the campus interactive text and the historical interactive text set, identifying the semantic continuation relationship between the campus interactive text and each historical interactive text in the historical interactive text set, and constructing a contextual semantic chain based on the semantic continuation relationship includes: Performing word segmentation processing on the text content of the campus interactive text to obtain a plurality of current text word units, and performing word segmentation processing on the input text content of each historical interactive text in the historical interactive text set to obtain a plurality of historical text word units; Extracting a core noun unit and a core verb unit from the current text word unit, wherein the core noun unit represents an object requested by the user, and the core verb unit represents an action requested by the user; Traversing each historical interactive text in the historical interactive text set, extracting core noun units and core verb units in the historical text word units of the historical interactive text, and matching them with the core noun units and core verb units of the current text word unit; If the core noun unit of any historical interactive text is the same as the core noun unit of the current text word unit, and the core verb unit is logically associated, then it is determined that the campus interactive text and the historical interactive text have a semantic continuity relationship, and the logical association includes an action continuation relationship, a state progression relationship, or an object association relationship; Historical interactive texts with semantic continuity relationships are sorted according to the interaction time information, and together with the campus interactive texts, they form a text sequence. The semantic dependency relationships between adjacent texts in the text sequence are analyzed, and a context semantic chain is constructed based on the text sequence and the semantic dependency relationships. The semantic dependency relationships include object dependency, condition dependency, and result dependency. Each node in the context semantic chain corresponds to a text fragment, and the lines between nodes represent semantic dependency relationships.

4. The text semantic understanding and analysis method applied to the smart campus platform according to claim 3 is characterized in that: If the core noun unit of any historical interactive text is the same as the core noun unit of the current text word unit, and the core verb unit is logically associated, then determining that the campus interactive text has a semantic continuity relationship with the historical interactive text includes: Performing a string comparison on the core noun unit of the current text word unit and the core noun unit of any of the historical interactive texts; if the two are identical, determining that the noun matching is successful; If the noun is successfully matched, the core verb unit of the current text word unit and the core verb unit of the historical interactive text are extracted, and the semantic distance between the two is queried through a semantic association dictionary, where the semantic association dictionary stores the semantic association relationship between verbs and the corresponding association strength; If the semantic distance is less than a preset threshold, it is determined that the core verb unit has a logical association, and the type of the logical association is determined according to the association relationship in the semantic association dictionary; If the noun matching is successful and the core verb unit has a logical association, the noun matching results and the verb association results are combined to determine whether the campus interactive text has a semantic continuity relationship with the historical interactive text, and the interaction time information and semantic dependency relationship type of the historical interactive text are recorded.

5. The text semantic understanding and analysis method applied to the smart campus platform according to claim 1 is characterized in that: The method of locating the associated campus resource status information from the campus resource status library of the smart campus platform based on the context semantic chain and generating a resource matching result by comparing the text fragment in the context semantic chain with the description text of the campus resource status information includes: Parsing the context semantic chain to extract resource-pointing nouns in each text segment; Determining a target resource type based on the resource pointing noun; Access the corresponding sub-library of the campus resource status library of the smart campus platform according to the target resource type; Retrieving resource status records containing the resource pointing noun in the corresponding sub-library, extracting description text of each resource status record, and performing semantic similarity comparison between the description text and the input text content of the corresponding text segment in the context semantic chain; The resource status record with the highest semantic similarity is determined as the associated campus resource status information, and the status description field, person in charge information field and update time field in the campus resource status information are extracted and combined to generate a resource matching result.

6. The text semantic understanding and analysis method applied to the smart campus platform according to claim 1 is characterized in that: The receiving of the subsequent interactive text fed back by the user terminal, adding the subsequent interactive text to the context semantic chain and updating the semantic dependency relationship between each fragment, and adjusting the query parameter of the campus resource status library based on the updated context semantic chain, includes: monitoring, within a preset feedback time window, subsequent interactive text sent by the user terminal corresponding to the user identifier, wherein the subsequent interactive text is the user's response to the semantic understanding result, including a confirmation statement, a supplementary statement, or a correction statement; Performing word segmentation processing on the subsequent interactive text to obtain subsequent text word units, extracting core noun units and core verb units from the subsequent text word units, and determining a semantic dependency relationship between the subsequent interactive text and the last text segment in the context semantic chain based on the core noun units and core verb units; Adding the subsequent interactive text as a new node to the end of the context semantic chain, recording the interaction time information of the node as the time of receiving the subsequent interactive text, and establishing a semantic dependency relationship between the new node and the previous node; Traverse the updated context semantic chain and recalculate the strength of the semantic dependency between nodes; Based on the resource-pointing nouns and semantic dependency strength in the updated context semantic chain, the query parameters of the campus resource status library are adjusted, the retrieval weight of resource-pointing nouns corresponding to high dependency strength is increased, and the retrieval scope of relevant resource status records is expanded to optimize the accuracy of resource positioning in the next semantic association processing.

7. A text semantic understanding and analysis system applied to a smart campus platform, characterized in that: The method comprises a processor and a machine-readable storage medium, wherein the machine-readable storage medium is connected to the processor, the machine-readable storage medium is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the machine-readable storage medium to implement the method described in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that It is used to store programs, instructions or codes, and when the programs, instructions or codes are executed by a processor, the method described in any one of claims 1 to 6 is implemented.

Images