Community management method and system based on artificial intelligence

Abstract

The invention relates to the technical field of community management, in particular to a community management method and system based on artificial intelligence. The method comprises the following steps: obtaining an owner question content; analyzing the question content of the owner, and determining a question item; based on the question item, utilizing artificial intelligence to obtain item data; summarizing project bills according to the project data; and checking the project bill to obtain and display a final project bill. The problems of delay and fragmentation of manual association of data sources are eliminated, the integrity and consistency of project data are ensured, data missing or conflicts are reduced, and traceability is enhanced. The problems of error proneness and static account division depending on manual summarization are avoided, the accuracy and dynamic adaptability of bills are improved, and the error rate is reduced. The project bill is checked, and the final project bill is obtained and displayed, so that the problems of omission and low efficiency of manual checking are avoided, the accuracy and the displayability of the final project bill are ensured, real-time and safe publicity is realized, the transparency is improved, and the owner trust is enhanced.

Description

Technical Field

[0001] This application relates to the field of community management technology, and in particular to a community management method and system based on artificial intelligence. Background Technology

[0002] In community management, fee management is a core aspect, involving the transparent and efficient handling of public revenues and expenditures such as property management fees, parking fees, and maintenance funds. These fees directly affect the quality of community services and residents' trust.

[0003] However, the existing community fee management system mainly relies on manual recording and public disclosure, resulting in low property fee collection rates, frequent parking fee arrears and false accounting, which in turn leads to a crisis of trust among residents and a decline in service quality. Summary of the Invention

[0004] This application provides a community management method and system based on artificial intelligence to solve the above problems.

[0005] Firstly, this application provides a community management method based on artificial intelligence, the method comprising:

[0006] Obtain the content of the homeowners' complaints; analyze the content of the homeowners' complaints to determine the items in question;

[0007] Based on the aforementioned questioned project, artificial intelligence was used to obtain project data;

[0008] Based on the project data, compile a project invoice;

[0009] Proofread the project invoices to obtain the final project invoices and display them.

[0010] This solution aims to: 1) Obtain homeowner complaints, improving the efficiency of complaint handling and reducing response delays caused by manual data entry; 2) Analyze homeowner complaints to identify the complained-about items, avoiding the time-consuming nature of manual analysis and its reliance on human experience, thus improving accuracy and efficiency and reducing the risk of overlooking semantic relationships; 3) Utilize artificial intelligence to acquire project data based on the complained-about items, eliminating the delays and fragmentation issues of manually linking data sources, ensuring the integrity and consistency of project data, reducing data loss or conflicts, and enhancing traceability; 4) Summarize project invoices based on project data, avoiding the error-proneness and static accounting issues of manual summarization, improving the accuracy and dynamic adaptability of invoices, and reducing the error rate; 5) Proofread project invoices to obtain and display the final project invoices, avoiding omissions and inefficiencies of manual proofreading, ensuring the accuracy and displayability of the final project invoices, achieving real-time and secure disclosure, improving transparency, and enhancing homeowner trust.

[0011] Optionally, the step of acquiring project data based on the questioned project using artificial intelligence includes:

[0012] Analyze the questioned items and determine the source of the project data;

[0013] Based on the project data source, the artificial intelligence is used to filter data content in a preset database to obtain project data.

[0014] This solution analyzes the challenged projects, identifies the sources of project data, eliminates the inability to automatically link multiple data sources, reduces redundant operations, and improves data retrieval efficiency. Based on the project data sources, artificial intelligence is used to filter data content in a pre-defined database to obtain project data, ensuring a strong correlation between the data and the challenge, avoiding omissions and errors from manual screening, and improving the accuracy and completeness of data acquisition.

[0015] Optionally, the step of parsing the homeowner's objection and determining the objection items includes:

[0016] Analyze the content of the homeowner's complaints and identify the keywords related to the complaints;

[0017] Based on a predefined type library, the questioning keywords are parsed to determine the keyword type;

[0018] Based on the aforementioned keyword types, keyword comparison criteria are determined;

[0019] Based on the keyword comparison criteria, determine the degree of relevance between any two questioning keywords;

[0020] Based on the aforementioned correlation, the items to be questioned are identified.

[0021] This solution analyzes homeowner complaints, identifies key terms, and reduces human intervention and errors. Based on a predefined type library, it analyzes complaint keywords, determines keyword types, enhances the accuracy of complaint analysis, ensures correct keyword classification, and avoids processing deviations due to type confusion. Based on keyword types, it establishes keyword comparison standards, avoiding subjectivity based on human experience and improving consistency and efficiency. According to the keyword comparison standards, it determines the correlation between any two complaint keywords, revealing potential connections and reducing fragmented analysis. Based on the correlation, it identifies the complaint items, transforming fragmented semantics into structured anomalies.

[0022] Optionally, the step of parsing the questioning keywords and determining the keyword type based on a predefined type library includes:

[0023] Parse the predefined type library to determine the preset part-of-speech tags;

[0024] Based on the preset part-of-speech tags, the questioning keywords are parsed to determine the initial type;

[0025] Based on the contextual position of the questioning keywords in the text, determine the semantic dependency relationship;

[0026] The keyword type is determined based on the initial type and the context correction factor.

[0027] This solution analyzes a predefined type library to determine preset part-of-speech tags, avoiding the risk of inconsistent categories encountered during manual operations. Based on the preset part-of-speech tags, it analyzes the questioning keywords to determine the initial type, reducing reliance on human experience and improving analysis efficiency. According to the contextual position of the questioning keywords in the text, it determines semantic dependencies, revealing the deeper meaning of the homeowner's questions and preventing the omission of details. Based on the initial type and contextual correction factors, it determines the keyword type, ensuring that the classification results more accurately reflect the semantic context of the homeowner's questions and improving the precision of keyword type matching.

[0028] Optionally, determining the questioned items based on the correlation includes:

[0029] Based on the aforementioned relevance, determine the degree centrality of the questioning keywords;

[0030] Based on the degree centrality, determine the core keyword group;

[0031] Based on the preset part-of-speech tags, the core keyword groups are analyzed to identify the items to be questioned.

[0032] This solution determines the degree centrality of the questioning keywords based on relevance, avoiding subjective biases inherent in manual parsing. Based on degree centrality, it identifies core keyword groups, enabling semantic focus on the questioning text and filtering out less important keywords, thereby simplifying the analysis scope and improving efficiency. By analyzing the core keyword groups using pre-defined part-of-speech tags, it identifies the questioning items, eliminating the problem of automatically failing to identify them.

[0033] Optionally, summarizing the project bill based on the project data includes:

[0034] Based on the questioned project, the project data was analyzed to determine the income and expenditure records and contract records;

[0035] Based on the contract records, the income and expenditure records are parsed to generate a traceable fund path;

[0036] Summarize project invoices based on the traceable funding path.

[0037] This solution analyzes project data based on the questioned project, identifying income and expenditure records and contract records. This avoids issues with the inability to automatically link data sources such as contracts and income / expense records, reducing manual analysis time and improving response efficiency. Based on contract records, income and expenditure records are analyzed to generate traceable fund paths, eliminating the problem of untraceable expense flows and enhancing the credibility of invoices. Based on traceable fund paths, project invoices are aggregated, eliminating the lack of invoice aggregation mechanisms and the inability to generate traceable project invoices, thereby increasing resident trust.

[0038] Optionally, the step of parsing the income and expenditure records based on the contract records to generate a traceable fund path includes:

[0039] Analyze the contract records to determine the text of the revenue sharing terms;

[0040] Analyze the revenue-sharing terms and conditions to determine the set of revenue-sharing participants;

[0041] Parse the income and expenditure records to determine the transaction fields;

[0042] The transaction fields are analyzed to determine the actual flow of funds and the timing of the fund flows.

[0043] Based on the time of fund flow, a traceable fund path is generated according to the set of revenue sharing participants and the actual fund flow.

[0044] This solution analyzes contract records to determine the text of revenue-sharing terms, avoiding interference from irrelevant clauses. It analyzes these terms to identify the set of participants, ensuring that the fund flow path only includes the parties stipulated in the contract. It analyzes income and expenditure records to identify transaction fields, ensuring that each transaction is processed individually. It analyzes these transaction fields to determine the actual fund flow and timing, ensuring the traceability of fund flow details. Based on the fund flow timing, and according to the set of participants and the actual fund flow, it generates a traceable fund path, enabling visualization and auditing support for fund flows.

[0045] Optionally, summarizing project invoices based on the traceable funding path includes:

[0046] Deconstruct the traceable funding path and identify key funding nodes;

[0047] Generate a node bill based on the income and expenditure records and the contract records;

[0048] Based on the time of fund flow, and according to the set of revenue sharing participants and the actual fund flow, the fund conflict node is determined;

[0049] Summarize the project bill based on the node bills and the funding conflict nodes.

[0050] This solution decomposes traceable funding paths, identifies key funding nodes, and ensures that the fund flow process is broken down into manageable units, avoiding interference from the overall complexity of the traceable funding path in data processing. Based on income and expenditure records and contract records, node bills are generated, ensuring that bill entries can be independently verified and audited. Based on the timeline of fund flows, and according to the set of participating parties and the actual flow of funds, conflict nodes are identified, providing a focus for project bill reconciliation and ensuring that potential errors are clearly marked. Based on the node bills and conflict nodes, project bills are summarized, eliminating the problems of high billing error rates and weak audit credibility.

[0051] Optionally, the step of verifying the project invoice, obtaining the final project invoice, and displaying it includes:

[0052] Based on the aforementioned funding conflict points, identify the conflicting projects;

[0053] Analyze the conflicting items to determine the conflicting amounts and details;

[0054] The level of abnormality is determined based on the amount of the conflict and the content of the conflict.

[0055] Based on the anomaly level, determine the handling method for the conflicting item, and based on the handling method, obtain and display the final item bill.

[0056] This solution identifies conflicting projects based on the points of financial conflict, avoiding full-scale verification and improving verification efficiency. It analyzes conflicting projects to determine the conflicting amounts and content, quantifying their scale and nature to make them more specific and assessable. Based on the conflicting amounts and content, it determines the anomaly level, classifying and grading conflicting projects to provide a basis for selecting handling methods and optimizing resource allocation. Based on the anomaly level, it determines the handling methods for each conflicting project, and based on these methods, obtains and displays the final project invoice, ensuring that conflicting projects are handled specifically, avoiding a one-size-fits-all approach, improving processing efficiency, and enhancing transparency and traceability.

[0057] Secondly, this application provides an artificial intelligence-based community management system, the system comprising:

[0058] The objection analysis module is used to obtain the content of the objections raised by homeowners; analyze the content of the objections raised by homeowners to determine the objection items;

[0059] The data acquisition module is used to acquire project data based on the questioned project using artificial intelligence.

[0060] The bill summary module is used to summarize project bills based on the project data;

[0061] The bill verification module is used to verify the project bills, obtain the final project bills, and display them.

[0062] Optionally, when the data acquisition module acquires project data based on the questioned project using artificial intelligence, it is used for:

[0063] Analyze the questioned items and determine the source of the project data;

[0064] Based on the project data source, the artificial intelligence is used to filter data content in a preset database to obtain project data.

[0065] Optionally, when the objection parsing module parses the owner's objection and determines the objection item, it is used for:

[0066] Analyze the content of the homeowner's complaints and identify the keywords related to the complaints;

[0067] Based on a predefined type library, the questioning keywords are parsed to determine the keyword type;

[0068] Based on the aforementioned keyword types, keyword comparison criteria are determined;

[0069] Based on the keyword comparison criteria, determine the degree of relevance between any two questioning keywords;

[0070] Based on the aforementioned correlation, the items to be questioned are identified.

[0071] Optionally, the questioning parsing module, based on a predefined type library, parses the questioning keywords and, when determining the keyword type, is used for:

[0072] Parse the predefined type library to determine the preset part-of-speech tags;

[0073] Based on the preset part-of-speech tags, the questioning keywords are parsed to determine the initial type;

[0074] Based on the contextual position of the questioning keywords in the text, determine the semantic dependency relationship;

[0075] The keyword type is determined based on the initial type and the context correction factor.

[0076] Optionally, when the questioning analysis module determines the questioned item based on the correlation, it is used to:

[0077] Based on the aforementioned relevance, determine the degree centrality of the questioning keywords;

[0078] Based on the degree centrality, determine the core keyword group;

[0079] Based on the preset part-of-speech tags, the core keyword groups are analyzed to identify the items to be questioned.

[0080] Optionally, when the bill summary module summarizes project bills based on the project data, it is used to:

[0081] Based on the questioned project, the project data was analyzed to determine the income and expenditure records and contract records;

[0082] Based on the contract records, the income and expenditure records are parsed to generate a traceable fund path;

[0083] Summarize project invoices based on the traceable funding path.

[0084] Optionally, when the bill summary module parses the income and expenditure records based on the contract records to generate a traceable fund path, it is used for:

[0085] Analyze the contract records to determine the text of the revenue sharing terms;

[0086] Analyze the revenue-sharing terms and conditions to determine the set of revenue-sharing participants;

[0087] Parse the income and expenditure records to determine the transaction fields;

[0088] The transaction fields are analyzed to determine the actual flow of funds and the timing of the fund flows.

[0089] Based on the time of fund flow, a traceable fund path is generated according to the set of revenue sharing participants and the actual fund flow.

[0090] Optionally, when the bill summary module summarizes project bills based on the traceable funding path, it is used for:

[0091] Deconstruct the traceable funding path and identify key funding nodes;

[0092] Generate a node bill based on the income and expenditure records and the contract records;

[0093] Based on the time of fund flow, and according to the set of revenue sharing participants and the actual fund flow, the fund conflict node is determined;

[0094] Summarize the project bill based on the node bills and the funding conflict nodes.

[0095] Optionally, when the bill verification module verifies the project bill to obtain and display the final project bill, it is used for:

[0096] Based on the aforementioned funding conflict points, identify the conflicting projects;

[0097] Analyze the conflicting items to determine the conflicting amounts and details;

[0098] The level of abnormality is determined based on the amount of the conflict and the content of the conflict.

[0099] Based on the anomaly level, determine the handling method for the conflicting item, and based on the handling method, obtain and display the final item bill. Attached Figure Description

[0100] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0101] Figure 1 This is a schematic diagram illustrating an application scenario provided in one embodiment of this application;

[0102] Figure 2 A flowchart illustrating an artificial intelligence-based community management method provided in one embodiment of this application;

[0103] Figure 3 This is a schematic diagram of the structure of an artificial intelligence-based community management system provided in one embodiment of this application. Detailed Implementation

[0104] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0105] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0106] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0107] The existing community fee management system mainly relies on manual recording and public disclosure, resulting in low property fee collection rates, frequent parking fee arrears and false accounting, which in turn leads to a crisis of trust among residents and a decline in service quality.

[0108] Based on this, this application provides an artificial intelligence-based community management method and system. This method acquires homeowner complaints, improving the efficiency of complaint processing and reducing response delays caused by manual data entry. It analyzes the homeowner complaints to identify the complaint items, avoiding the time-consuming nature of manual analysis and its reliance on human experience, thus improving accuracy and efficiency and reducing the risk of overlooking semantic relationships. Based on the complaint items, it utilizes artificial intelligence to acquire project data, eliminating the delays and fragmentation issues associated with manually linking data sources, ensuring the integrity and consistency of project data, reducing data loss or conflicts, and enhancing traceability. Based on the project data, it summarizes project bills, avoiding the error-proneness and static accounting issues of manual summarization, improving the accuracy and dynamic adaptability of bills, and reducing the error rate. Finally, it verifies the project bills to obtain and display the final project bills, avoiding omissions and inefficiencies of manual verification, ensuring the accuracy and displayability of the final project bills, achieving real-time and secure public disclosure, improving transparency, and enhancing homeowner trust.

[0109] Figure 1 This is a schematic diagram illustrating an application scenario provided by this application, showing the application of the method provided in this application during community management.

[0110] Specifically, the method provided in this application is applied to any server, where the server interacts with the user device to obtain the owner's query content. The query content is analyzed to determine the query item. Based on the query item, artificial intelligence is used to acquire project data, eliminating the latency and fragmentation issues of manually linking data sources, ensuring the integrity and consistency of project data, reducing data loss or conflicts, and enhancing traceability. Based on the project data, project invoices are summarized, avoiding the error-proneness and static accounting problems of relying on manual summarization, improving the accuracy and dynamic adaptability of invoices, and reducing the error rate. The project invoices are verified to obtain and display the final project invoice, avoiding the omissions and inefficiencies of manual verification, ensuring the accuracy and displayability of the final project invoice, achieving real-time and secure disclosure, improving transparency, and enhancing owner trust.

[0111] For specific implementation details, please refer to the following examples.

[0112] Figure 2 This is a flowchart illustrating an artificial intelligence-based community management method according to an embodiment of this application. The method of this embodiment can be applied to servers in the above scenarios. Figure 2 As shown, the method includes:

[0113] S201. Obtain the content of the owner's objection; analyze the content of the owner's objection and determine the objection items;

[0114] The homeowner's objection can be in the form of a text document submitted by the homeowner.

[0115] The project under question can be any project that the owner believes has abnormal fees.

[0116] Specifically, the server connects to user devices via an API interface; then, it retrieves information uploaded by users—specifically, the homeowners' complaints—from the user devices in real time through the API interface. Next, it uses natural language processing technology to segment and identify entities from the homeowners' complaints, extracting keywords. These keywords are then mapped to a predefined fee item library within a standard classification system for community fee management. String matching is used to compare the keywords with entries in the predefined fee item library. When a keyword successfully matches a specific entry in the predefined fee item library, the complaint item is identified.

[0117] S202. Based on the questioning of the project, use artificial intelligence to obtain project data;

[0118] Artificial intelligence can be an automated technology module used to analyze questions, obtain project data, and verify invoices.

[0119] Project data can be structured data related to the project being questioned.

[0120] Specifically, a query statement is constructed based on the questioned project; then, artificial intelligence is used to filter data such as contract terms, historical maintenance work orders, and actual income and expenditure records that match the query statement, i.e., project data.

[0121] S203. Summarize project invoices based on project data;

[0122] Project invoices can be a summary report of expenses.

[0123] Specifically, artificial intelligence applications utilize natural language processing to analyze multiple items (such as transaction records or fee categories) in project data. Multiple items are calculated through iterative algorithms, grouped, and summarized to generate a project invoice.

[0124] S204. Proofread the project invoice, obtain the final project invoice, and display it.

[0125] The final project invoice can be a proofread and displayable invoice.

[0126] Specifically, artificial intelligence is used to proofread project bills, check their consistency, and identify inconsistencies, i.e., potential conflicts, such as timestamp conflicts (transaction time earlier than contract effective time) and amount conflicts (property fee allocation calculation error leading to mismatch in total amount).

[0127] The system uses artificial intelligence to automatically extract frequently occurring billing errors from historical billing error records that failed manual or automatic verification in community management. Based on these errors, historical audit rules are generated. Each identified error is then matched against the rule conditions to check for rule violations. Errors are adjusted, annotations or tags are added, and corrections are made. All data is then integrated to generate the final project bill. Finally, the final project bill is displayed through a visual interface.

[0128] This solution aims to: 1) Obtain homeowner complaints, improving the efficiency of complaint handling and reducing response delays caused by manual data entry; 2) Analyze homeowner complaints to identify the complained-about items, avoiding the time-consuming nature of manual analysis and its reliance on human experience, thus improving accuracy and efficiency and reducing the risk of overlooking semantic relationships; 3) Utilize artificial intelligence to acquire project data based on the complained-about items, eliminating the delays and fragmentation issues of manually linking data sources, ensuring the integrity and consistency of project data, reducing data loss or conflicts, and enhancing traceability; 4) Summarize project invoices based on project data, avoiding the error-proneness and static accounting issues of manual summarization, improving the accuracy and dynamic adaptability of invoices, and reducing the error rate; 5) Proofread project invoices to obtain and display the final project invoices, avoiding omissions and inefficiencies of manual proofreading, ensuring the accuracy and displayability of the final project invoices, achieving real-time and secure disclosure, improving transparency, and enhancing homeowner trust.

[0129] In some embodiments, the questioned project is analyzed to determine the source of the project data; based on the source of the project data, artificial intelligence is used to filter the data content in a preset database to obtain the project data.

[0130] The source of project data can be the project data storage location.

[0131] The default database can be a database for storing community fee information such as contract records and transaction details. It is pre-stored on the server and accessed when needed.

[0132] The data content can be relevant structured data content selected from a preset database.

[0133] Specifically, based on the questioned project, natural language processing technology is used to break down the questioned content into word or phrase units through word segmentation. Then, entities (such as time entities, amount entities, or fee category entities) are extracted from the word segmentation results through scanning and classification. Furthermore, based on the extracted entities, a preset rule base for community fee scenarios manually collected by the storage administrator is used for string matching (each rule is associated with a project data source). The entities are compared with the entries in the preset rule base to determine the project data source.

[0134] Based on the project data source, query conditions are generated. Artificial intelligence is used to call the database API to transmit the query conditions in real time. The query conditions are then used to filter the data content in a preset database that stores community fee management data based on a manually created and maintained simple database, excluding irrelevant entries, thereby obtaining the project data.

[0135] This solution analyzes the challenged projects, identifies the sources of project data, eliminates the inability to automatically link multiple data sources, reduces redundant operations, and improves data retrieval efficiency. Based on the project data sources, artificial intelligence is used to filter data content in a pre-defined database to obtain project data, ensuring a strong correlation between the data and the challenge, avoiding omissions and errors from manual screening, and improving the accuracy and completeness of data acquisition.

[0136] In some embodiments, the content of the owner's objection is parsed to determine the objection keywords; based on a predefined type library, the objection keywords are parsed to determine the keyword type; based on the keyword type, the keyword comparison standard is determined; according to the keyword comparison standard, the relevance between any two objection keywords is determined; and based on the relevance, the objection item is determined.

[0137] The keywords for questioning can be words or phrases extracted from the content of the homeowner's question.

[0138] A predefined type library can be a pre-defined category library containing fixed categories such as cost types, exception types, and entity types.

[0139] The keyword type can be the category to which the keyword belongs after matching a predefined type library.

[0140] Keyword comparison criteria can be based on comparison rules set according to keyword type.

[0141] Relevance can be a measure of the strength of the correlation between any two questioning keywords.

[0142] Specifically, based on the content of the homeowner's complaints, natural language processing technology is used to segment the content into words. By filtering out non-key units, the keywords for the complaints are determined. Then, each keyword is traversed and matched against a predefined type library that maps storage time entities, amount entities, and expense category entities to keyword types based on taxonomy theory and manually defined standard categories. A corresponding keyword type is then assigned to each keyword.

[0143] Based on keyword type, logical conditions are extracted from preset rules in a database that stores the mapping relationship between keyword type and logical conditions to determine the corresponding keyword comparison criteria. Then, based on these criteria, any two questioning keywords are traversed, and a rule engine with preset rules is used to calculate the relevance between them. Finally, an aggregation method is used to summarize the relevance, and preset aggregation rules based on data aggregation theory are applied to filter highly relevant questioning keyword pairs; subsequently, the questioned items are determined based on these highly relevant questioning keyword pairs.

[0144] This solution analyzes homeowner complaints, identifies key terms, and reduces human intervention and errors. Based on a predefined type library, it analyzes complaint keywords, determines keyword types, enhances the accuracy of complaint analysis, ensures correct keyword classification, and avoids processing deviations due to type confusion. Based on keyword types, it establishes keyword comparison standards, avoiding subjectivity based on human experience and improving consistency and efficiency. According to the keyword comparison standards, it determines the correlation between any two complaint keywords, revealing potential connections and reducing fragmented analysis. Based on the correlation, it identifies the complaint items, transforming fragmented semantics into structured anomalies.

[0145] In some embodiments, a predefined type library is parsed to determine preset part-of-speech tags; based on the preset part-of-speech tags, questioning keywords are parsed to determine the initial type; based on the contextual position of the questioning keywords in the text, semantic dependencies are determined; and based on the initial type and contextual correction factor, the keyword type is determined.

[0146] Predefined part-of-speech tags can be tag categories parsed from a predefined type library, representing the syntactic or semantic categories of keywords. They are pre-stored on the server and invoked when needed.

[0147] The initial type can be a preliminary type of questioning keywords.

[0148] Contextual position can be used to question the location of keywords in the text.

[0149] Semantic dependency can be used to question the grammatical dependency between keywords and neighboring words.

[0150] Context correction factors can be adjustment factors generated based on semantic dependencies, which can correct for biases in the initial type.

[0151] Specifically, the system iterates through the entries of each category in the predefined type library and extracts several category names as preset part-of-speech tags. Then, it performs string matching between the questioning keyword and each tag in the preset part-of-speech tags, checking whether the questioning keyword is completely equal to the preset part-of-speech tag, or whether the questioning keyword contains a substring of the preset part-of-speech tag. If the questioning keyword matches or partially contains the tag entry, the match is successful, and the corresponding tag is assigned as the initial type.

[0152] The standard string search algorithm is used to locate the contextual position of the questioning keyword in the text. Then, the dependency parsing algorithm is used to segment the text and tag its parts of speech. After segmentation and tagging, the dependency parsing algorithm is applied to construct a syntax tree. Based on the grammatical dependency rules established according to the dependency relation types defined by general dependency parsing (such as subject-verb relation, object relation), the semantic dependency relations between the questioning keyword and neighboring words are scanned in the syntax tree. For example, if the rule is "noun modifies adjective → modification relation", then when the questioning keyword is a noun and the neighboring word is an adjective, it is identified as a "modification relation".

[0153] Context correction factors are generated based on semantic dependency relationships. Finally, the initial type and context correction factors are combined, and the condition-action rule engine is applied for mapping. The condition-action rules are traversed, and the keyword type is determined when the conditions are completely matched.

[0154] This solution analyzes a predefined type library to determine preset part-of-speech tags, avoiding the risk of inconsistent categories encountered during manual operations. Based on the preset part-of-speech tags, it analyzes the questioning keywords to determine the initial type, reducing reliance on human experience and improving analysis efficiency. According to the contextual position of the questioning keywords in the text, it determines semantic dependencies, revealing the deeper meaning of the homeowner's questions and preventing the omission of details. Based on the initial type and contextual correction factors, it determines the keyword type, ensuring that the classification results more accurately reflect the semantic context of the homeowner's questions and improving the precision of keyword type matching.

[0155] In some embodiments, the degree centrality of the questioned keywords is determined based on relevance; the core keyword group is determined based on the degree centrality; and the core keyword group is analyzed based on preset part-of-speech tags to determine the questioned items.

[0156] Degree centrality can be used as an indicator to question the centrality of a keyword in a keyword relationship graph.

[0157] The core keyword group can be a subset of highly important keywords selected from those that raise questions.

[0158] Specifically, based on relevance, a keyword relationship graph is constructed: each questioning keyword is treated as a node, and the relevance is used as the edge weight between nodes.

[0159] For each questioning keyword node, the degree centrality of the questioning keyword is determined by summing the relevance of all edge weights connected to that node. Then, the questioning keywords are sorted in descending order of degree centrality. Based on the sorting results, a core keyword group with a degree centrality higher than the preset threshold (based on human experience) is selected. Next, based on the core keyword group, preset part-of-speech tags are matched using string matching. The keyword string is compared with the tag string (e.g., the keyword "property fee" is compared with the tag "fee entity"; if they are equal, they match), or it is checked whether the keyword contains a subsequence of the tag (e.g., the keyword "service fee" contains a substring of the tag "fee," then it matches). Then, based on the matching results, the frequency of preset part-of-speech tags for several core keywords in the core keyword group is counted. Finally, the preset part-of-speech tag with the highest frequency is selected as the questioning item.

[0160] This solution determines the degree centrality of the questioning keywords based on relevance, avoiding subjective biases inherent in manual parsing. Based on degree centrality, it identifies core keyword groups, enabling semantic focus on the questioning text and filtering out less important keywords, thereby simplifying the analysis scope and improving efficiency. By analyzing the core keyword groups using pre-defined part-of-speech tags, it identifies the questioning items, eliminating the problem of automatically failing to identify them.

[0161] In some embodiments, based on the questioned project, project data is parsed to determine income and expenditure records and contract records; based on the contract records, income and expenditure records are parsed to generate traceable funding paths; based on the traceable funding paths, project bills are summarized.

[0162] Income and expenditure records can be the transaction flow of expenses and expenditures in project data. Income and expenditure records include core information such as fund transfers in / out, amount, and time.

[0163] Contract records can include contract terms and agreements related to fees.

[0164] A traceable funding path can be the complete path of fund flow, clearly showing the trajectory of funds from source to destination.

[0165] Specifically, based on the project under investigation, keyword matching is used to retrieve income and expenditure records related to the project from the project data; at the same time, based on the contract identifier associated with the project under investigation, the corresponding contract records are retrieved from the project data.

[0166] Natural language processing (NLP) technology is used to analyze contract records and extract key constraints (a set of rules parsed from the contract records). These key constraints include monetary constraints for payments or receipts, time constraints for deadlines or dates, conditions for additional clauses (such as payment upon acceptance), and constraints on participating entities (such as Party A and Party B) to ensure the compliance of fund flows. Then, based on these key constraints, each income and expenditure record is compared with the contract record to determine if it meets the constraints. For each record that meets the constraints, a path node is created (the path node includes the transaction ID, amount, and time). Linking these path nodes generates a traceable fund path.

[0167] The system iterates through each transaction record in the traceable funding path, extracts fund flow data (information units extracted from transaction records), and describes the core attributes of a single fund flow. Fund flow data includes a unique transaction identifier, positive numbers representing income and negative numbers representing expenditure, a timestamp indicating the transaction date, the participating parties (source account and target account), and the transaction type. Then, an aggregation algorithm is applied to read the total income and expenditure and details of fund flow data node by node, thereby generating a project invoice.

[0168] This solution analyzes project data based on the questioned project, identifying income and expenditure records and contract records. This avoids issues with the inability to automatically link data sources such as contracts and income / expense records, reducing manual analysis time and improving response efficiency. Based on contract records, income and expenditure records are analyzed to generate traceable fund paths, eliminating the problem of untraceable expense flows and enhancing the credibility of invoices. Based on traceable fund paths, project invoices are aggregated, eliminating the lack of invoice aggregation mechanisms and the inability to generate traceable project invoices, thereby increasing resident trust.

[0169] In some embodiments, contract records are parsed to determine the revenue sharing terms; the revenue sharing terms are analyzed to determine the set of revenue sharing participants; income and expenditure records are parsed to determine transaction fields; transaction fields are parsed to determine the actual flow of funds and the time of fund flow; based on the time of fund flow, a traceable fund path is generated according to the set of revenue sharing participants and the actual flow of funds.

[0170] The revenue sharing clause can be the content of the clauses involving fund sharing in the contract record. The revenue sharing clause includes core constraints such as the revenue sharing ratio and the definition of the participating parties.

[0171] The set of revenue sharing participants can be the set of entities that share the funds as specified in the revenue sharing terms.

[0172] Transaction fields can be fields in income and expense records that describe a single transaction, and these fields include details such as the transacting party and the amount.

[0173] The actual flow of funds can be the actual source and destination of funds in a transaction.

[0174] The timing of fund flows can be the point in time when a transaction occurs.

[0175] Specifically, the text content of the contract record is matched with keywords to identify clauses related to revenue sharing; the matched clauses are then used as revenue sharing clause text. Subsequently, the revenue sharing clause text is scanned word by word to extract nouns representing the parties sharing the funds (the entities agreed upon in the contract to share the funds); then, the extracted nouns are deduplicated and combined into a set of revenue sharing participants.

[0176] Each transaction entry in the income and expenditure records is traversed, and the transaction fields are extracted. Then, the transaction party string in each transaction field is parsed to identify the source and destination, determining the actual flow of funds. Simultaneously, the timestamp in each transaction field is read to determine the time of fund flow. Next, the actual fund flows are filtered, checking whether the source or destination is in the set of profit-sharing participants, retaining transactions involving entities in the profit-sharing participant set. Finally, based on the fund flow time, the filtered transactions are sorted in ascending order of time to form a time series. The time-series sorted transaction list represents the traceable fund path.

[0177] This solution analyzes contract records to determine the text of revenue-sharing terms, avoiding interference from irrelevant clauses. It analyzes these terms to identify the set of participants, ensuring that the fund flow path only includes the parties stipulated in the contract. It analyzes income and expenditure records to identify transaction fields, ensuring that each transaction is processed individually. It analyzes these transaction fields to determine the actual fund flow and timing, ensuring the traceability of fund flow details. Based on the fund flow timing, and according to the set of participants and the actual fund flow, it generates a traceable fund path, enabling visualization and auditing support for fund flows.

[0178] In some embodiments, the traceable funding path is decomposed to identify key funding nodes; node bills are generated based on income and expenditure records and contract records; funding conflict nodes are identified based on the time of fund flow, the set of revenue sharing participants, and the actual fund flow; and project bills are summarized based on node bills and funding conflict nodes.

[0179] Key funding nodes can be funding flow event units, consisting of a tripartite combination of the time, source, and destination of a single transaction.

[0180] A node bill can be a bill entry generated for each key funding node. The node bill contains transaction data (amount, time, and transacting party) and revenue sharing rules.

[0181] A funding conflict node can be a conflict point that exists in a key funding node. A funding conflict node includes the conflict type (time violation, participant violation, sharing violation) and the associated node bill.

[0182] Specifically, each transaction event in the traceable funding path is traversed. Then, for each transaction event, the time, source, and destination fields are extracted and marked as key funding nodes. Next, transaction events in the income and expenditure records that share the same time and trading parties as the key funding nodes are searched, and transaction fields are extracted. Simultaneously, the profit-sharing clause text of the contract record is parsed, and date keywords in the profit-sharing clause text are scanned to extract the time range and determine time constraints. Proportion keywords (such as allocation, ratio) in the profit-sharing clause text are scanned to extract the allocation logic and generate allocation rules. This yields profit-sharing rules containing time constraints and allocation rules. Finally, combining the transaction fields and profit-sharing rules, a node bill is generated.

[0183] The process involves verifying whether the timing of fund flows conforms to the time constraints in the contract records; then, verifying whether the source and destination of the actual fund flow are both within the set of revenue sharing participants, and whether the actual fund flow conforms to the allocation rules in the revenue sharing terms; if any verification fails, it is marked as a fund conflict node. Next, several node invoices are merged to form an initial summary list; then, the fund conflict nodes are traversed, the corresponding node invoices are queried in the initial summary list, and conflict markers are added; finally, the node invoices with conflict markers are summarized to generate the project invoice.

[0184] This solution decomposes traceable funding paths, identifies key funding nodes, and ensures that the fund flow process is broken down into manageable units, avoiding interference from the overall complexity of the traceable funding path in data processing. Based on income and expenditure records and contract records, node bills are generated, ensuring that bill entries can be independently verified and audited. Based on the timeline of fund flows, and according to the set of participating parties and the actual flow of funds, conflict nodes are identified, providing a focus for project bill reconciliation and ensuring that potential errors are clearly marked. Based on the node bills and conflict nodes, project bills are summarized, eliminating the problems of high billing error rates and weak audit credibility.

[0185] In some embodiments, conflicting items are identified based on the funding conflict nodes; conflicting items are analyzed to determine the conflict amount and content; anomaly levels are determined based on the conflict amount and content; handling methods for conflicting items are determined based on the anomaly levels; and the final project bill is obtained and displayed based on the handling methods.

[0186] Conflicting projects can be identified and categorized from conflicting funding points.

[0187] The conflict amount can be the transaction amount involved in the conflicting project.

[0188] The content of the conflict can be descriptive of the conflicting items.

[0189] The level of anomaly can be determined based on the amount of the conflict and the severity of the conflict content.

[0190] The handling method can be a method specified based on the exception level.

[0191] Specifically, the process iterates through the conflict nodes, extracting the transaction event identifier (a unique string identifying a transaction event; to accurately associate conflict nodes with project bill entries, it is bound to the conflict node when storing the node bill) for each conflict node. The process then searches the project bill for a bill entry that exactly matches the transaction event identifier. Once a match is found, that bill entry is identified as the conflicting project. Subsequently, the amount field is read from the project bill corresponding to the conflicting project to determine the conflict amount. Simultaneously, based on the attributes of the conflict node (its metadata set, including conflict type, conflict details, and associated identifiers), and combined with the conflict amount, the conflict content is determined.

[0192] The application is manually configured based on human experience, using analysis of historical billing data (past project bills) and conflict cases (stored historical conflict events that have been processed, typically stored in a database and retrieved when needed) to set predefined rules (used to assess the anomaly level of conflicting projects). In the specific implementation, corresponding handling methods are established according to the conflict cases and the levels corresponding to the predefined rules.

[0193] At this point, based on the size of the conflict amount and the severity of the conflict content (the degree of harm of the financial conflict node, such as low severity (single, recoverable low severity), medium severity (recurring but controllable), and high severity (frequent participant violations), an anomaly level is determined according to predefined rules. For example, if the conflict amount is low but the severity is high (e.g., frequent participant conflicts), the handling measures are upgraded; if the conflict amount is high but the severity is low (e.g., single time deviation), the handling is downgraded. Furthermore, corresponding handling measures for each anomaly level are assigned: First, for low anomaly levels, the handling measure is automatic correction; for example, to correct time conflicts, the transaction time is adjusted to the most recent time point within the contract window; to correct allocation conflicts, the amount is recalculated according to the allocation rules and the original value is overwritten. Second, for medium anomaly levels, the handling measure is marked as pending review. Then, for high anomaly levels, the handling measure is marked as requiring manual intervention. Finally, the project bill is updated according to the handling measures, and the handling measures for several conflict items are integrated to generate the final project bill. The final project bill is then displayed through the interface.

[0194] This solution identifies conflicting projects based on the points of financial conflict, avoiding full-scale verification and improving verification efficiency. It analyzes conflicting projects to determine the conflicting amounts and content, quantifying their scale and nature to make them more specific and assessable. Based on the conflicting amounts and content, it determines the anomaly level, classifying and grading conflicting projects to provide a basis for selecting handling methods and optimizing resource allocation. Based on the anomaly level, it determines the handling methods for each conflicting project, and based on these methods, obtains and displays the final project invoice, ensuring that conflicting projects are handled specifically, avoiding a one-size-fits-all approach, improving processing efficiency, and enhancing transparency and traceability.

[0195] Figure 3 A schematic diagram of the structure of an artificial intelligence-based community management system provided in one embodiment of this application is shown below. Figure 3 As shown, the AI-based community management system 300 in this embodiment includes: a questioning and analysis module 301, a data acquisition module 302, a bill summary module 303, and a bill verification module 304.

[0196] The objection analysis module 301 is used to obtain the content of the owner's objection; analyze the content of the owner's objection, and determine the objection items;

[0197] The data acquisition module 302 is used to acquire project data based on the questioned project using artificial intelligence;

[0198] The bill summary module 303 is used to summarize project bills based on the project data;

[0199] The bill verification module 304 is used to verify the project bill, obtain the final project bill, and display it.

[0200] Optionally, when the data acquisition module 302 acquires project data based on the questioned project using artificial intelligence, it is used for:

[0201] Analyze the questioned items and determine the source of the project data;

[0202] Based on the project data source, the artificial intelligence is used to filter data content in a preset database to obtain project data.

[0203] Optionally, when the objection parsing module 301 parses the owner's objection and determines the objection item, it is used for:

[0204] Analyze the content of the homeowner's complaints and identify the keywords related to the complaints;

[0205] Based on a predefined type library, the questioning keywords are parsed to determine the keyword type;

[0206] Based on the aforementioned keyword types, keyword comparison criteria are determined;

[0207] Based on the keyword comparison criteria, determine the degree of relevance between any two questioning keywords;

[0208] Based on the aforementioned correlation, the items to be questioned are identified.

[0209] Optionally, the questioning parsing module 301, based on a predefined type library, parses the questioning keywords and determines the keyword type, for the following purposes:

[0210] Parse the predefined type library to determine the preset part-of-speech tags;

[0211] Based on the preset part-of-speech tags, the questioning keywords are parsed to determine the initial type;

[0212] Based on the contextual position of the questioning keywords in the text, determine the semantic dependency relationship;

[0213] The keyword type is determined based on the initial type and the context correction factor.

[0214] Optionally, when the questioning analysis module 301 determines the questioned item based on the correlation, it is used to:

[0215] Based on the aforementioned relevance, determine the degree centrality of the questioning keywords;

[0216] Based on the degree centrality, determine the core keyword group;

[0217] Based on the preset part-of-speech tags, the core keyword groups are analyzed to identify the items to be questioned.

[0218] Optionally, when the bill summary module 303 summarizes the project bills based on the project data, it is used for:

[0219] Based on the questioned project, the project data was analyzed to determine the income and expenditure records and contract records;

[0220] Based on the contract records, the income and expenditure records are parsed to generate a traceable fund path;

[0221] Summarize project invoices based on the traceable funding path.

[0222] Optionally, when the bill summary module 303 parses the income and expenditure records based on the contract records to generate a traceable fund path, it is used for:

[0223] Analyze the contract records to determine the text of the revenue sharing terms;

[0224] Analyze the revenue-sharing terms and conditions to determine the set of revenue-sharing participants;

[0225] Parse the income and expenditure records to determine the transaction fields;

[0226] The transaction fields are analyzed to determine the actual flow of funds and the timing of the fund flows.

[0227] Based on the time of fund flow, a traceable fund path is generated according to the set of revenue sharing participants and the actual fund flow.

[0228] Optionally, when the bill summary module 303 summarizes project bills based on the traceable funding path, it is used for:

[0229] Deconstruct the traceable funding path and identify key funding nodes;

[0230] Generate a node bill based on the income and expenditure records and the contract records;

[0231] Based on the time of fund flow, and according to the set of revenue sharing participants and the actual fund flow, the fund conflict node is determined;

[0232] Summarize the project bill based on the node bills and the funding conflict nodes.

[0233] Optionally, when the bill verification module 304 verifies the project bill to obtain and display the final project bill, it is used for:

[0234] Based on the aforementioned funding conflict points, identify the conflicting projects;

[0235] Analyze the conflicting items to determine the conflicting amounts and details;

[0236] The level of abnormality is determined based on the amount of the conflict and the content of the conflict.

[0237] Based on the anomaly level, determine the handling method for the conflicting item, and based on the handling method, obtain and display the final item bill.

[0238] The system in this embodiment can be used to execute the methods of any of the above embodiments, and its implementation principle and technical effect are similar, so they will not be described again here.

Claims

1. A community management method based on artificial intelligence, characterized in that, include: Obtain the content of the homeowner's questions; Analyze the content of the homeowner's objections and identify the items in question; Based on the aforementioned questioned project, artificial intelligence was used to obtain project data; Based on the project data, compile a project invoice; Proofread the project invoices to obtain the final project invoices and display them.

2. The method according to claim 1, characterized in that, The process of acquiring project data using artificial intelligence based on the questioned project includes: Analyze the questioned items and determine the source of the project data; Based on the project data source, the artificial intelligence is used to filter data content in a preset database to obtain project data.

3. The method according to claim 1, characterized in that, The process of analyzing the homeowner's complaints and identifying the items in question includes: Analyze the content of the homeowner's complaints and identify the keywords related to the complaints; Based on a predefined type library, the questioning keywords are parsed to determine the keyword type; Based on the aforementioned keyword types, determine the keyword comparison criteria; Based on the keyword comparison criteria, determine the degree of relevance between any two questioning keywords; Based on the aforementioned correlation, the items to be questioned are identified.

4. The method according to claim 3, characterized in that, The process of parsing the questioning keywords based on a predefined type library and determining the keyword type includes: Parse the predefined type library to determine the preset part-of-speech tags; Based on the preset part-of-speech tags, the questioning keywords are parsed to determine the initial type; Based on the contextual position of the questioning keywords in the text, determine the semantic dependency relationship; The keyword type is determined based on the initial type and the context correction factor.

5. The method according to claim 4, characterized in that, The process of determining the questionable items based on the correlation includes: Based on the aforementioned relevance, determine the degree centrality of the questioning keywords; Based on the degree centrality, determine the core keyword group; Based on the preset part-of-speech tags, the core keyword groups are analyzed to identify the items to be questioned.

6. The method according to claim 1, characterized in that, The step of summarizing project bills based on the project data includes: Based on the questioned project, the project data was analyzed to determine the income and expenditure records and contract records; Based on the contract records, the income and expenditure records are parsed to generate a traceable fund path; Summarize project invoices based on the traceable funding path.

7. The method according to claim 6, characterized in that, The step of parsing the income and expenditure records based on the contract records to generate a traceable fund path includes: Analyze the contract records to determine the text of the revenue sharing terms; Analyze the revenue-sharing terms and conditions to determine the set of revenue-sharing participants; Parse the income and expenditure records to determine the transaction fields; The transaction fields are analyzed to determine the actual flow of funds and the timing of the fund flows. Based on the time of fund flow, a traceable fund path is generated according to the set of revenue sharing participants and the actual fund flow.

8. The method according to claim 7, characterized in that, The process of summarizing project invoices based on the traceable funding path includes: Deconstruct the traceable funding path and identify key funding nodes; Generate a node bill based on the income and expenditure records and the contract records; Based on the time of fund flow, and according to the set of revenue sharing participants and the actual fund flow, the fund conflict node is determined; Summarize the project bill based on the node bills and the funding conflict nodes.

9. The method according to claim 8, characterized in that, The process of verifying the project invoice, obtaining the final project invoice, and displaying it includes: Based on the aforementioned funding conflict points, identify the conflicting projects; Analyze the conflicting items to determine the conflicting amounts and details; The level of abnormality is determined based on the amount of the conflict and the content of the conflict. Based on the anomaly level, determine the handling method for the conflicting item, and based on the handling method, obtain and display the final item bill.

10. A community management system based on artificial intelligence, characterized in that, The method applied to any one of claims 1-9 includes: The objection analysis module is used to obtain the content of the objections raised by homeowners; analyze the content of the objections raised by homeowners to determine the objection items; The data acquisition module is used to acquire project data based on the questioned project using artificial intelligence. The bill summary module is used to summarize project bills based on the project data; The bill verification module is used to verify the project bills, obtain the final project bills, and display them.

Images