72 results about "Opinion analysis" patented technology

The invention relates to the technical field of semantic recognition, in particular to a cross-border commodity multi-language comment semantic analysis method, which comprises the following steps of: acquiring a cross-border commodity multi-language comment text, constructing nodes and directed edges according to part-of-speech tagging and dependency relationship rules, generating a text syntax dependency directed graph, and generating a text syntax dependency directed graph; and traversing the text syntax dependency directed graph, calculating a node shortest dependency path between nodes, and generating a pruned syntax subtree. According to the method, the emotion consistency coefficient index is calculated and compared with the logic smoothness threshold value, so that the emotion stability and logic self-consistency in the comments can be quantitatively evaluated. The mechanism not only completes judgment of single-dimension emotional tendency, but also introduces verification of text logic quality, and effectively recognizes and filters out semantic contradictions and false comments or low-quality texts with obvious splicing traces, so that a cross-border commodity semantic analysis set which contains core semantic information and has high credibility is generated; and the precision of public opinion analysis is improved.

The invention provides a vehicle public opinion intelligent analysis method and system based on the Internet of Vehicles, a product and a vehicle, and relates to the technical field of vehicle public opinion analysis. The emotion contribution degree of each data source is dynamically evaluated and weighted by using a cross-modal attention mechanism, so that the one-sidedness of single-modal analysis is effectively overcome, and the accuracy and robustness of public opinion emotion judgment are improved; further deeply mining public opinion propagation information by means of a space-time diagram neural network, constructing a propagation path model, generating an event-emotion-propagation force three-dimensional thermodynamic diagram, and realizing visual insight and trend prediction of a public opinion evolution situation; on the basis, a local and external collaborative grading emergency response mechanism is constructed, user anxiety is relieved, and diffusion of bad public opinions is effectively restrained.

The invention discloses a public opinion data structuring method based on a Re-Act framework, and relates to the technical field of public opinion data processing and intelligent structuralization, and the method comprises the steps: firstly constructing a Re-Act public opinion structuring system comprising a thinking module, an action module and an observation module, then collecting data through multiple rounds of closed-loop iteration until a full link is covered, and finally obtaining the data. And finally, processing based on the full-link data through a structured tool, outputting standardized data and a relation graph, and terminating. According to the scheme, public opinion cause, response, action, interaction and ending full-link data are dynamically complemented through multiple rounds of closed loops of thinking, action and observation under the Re-Act framework, structured quality is optimized by combining multi-module processing, standardized data and a relation graph are output, public opinion dynamic characteristics are adapted, a complete and accurate support is provided for deep analysis and decision making, and the public opinion analysis and decision making efficiency is improved. And the limitation that the link is easy to break and the adaptability is poor in the traditional method is overcome.

The application relates to a construction method for obtaining a viewpoint evaluation object model, and belongs to the field of public opinion analysis. An event abstract text generation model is established, which is used for inputting information text containing event content to obtain event abstract text; a sentence annotation model is established, which is used for inputting viewpoint text for evaluating the event and the event abstract text to obtain a viewpoint with the event abstract text as semantic guidance; wherein the viewpoint is a sentence sequence extracted from the viewpoint text; a viewpoint evaluation object extraction model is established, which is used for inputting the viewpoint and the event abstract text to obtain an evaluation object of the viewpoint; wherein the evaluation object is a word sequence extracted from the event abstract text. The application can accurately obtain a viewpoint with the event as an evaluation object, and improves the analysis performance.

The invention belongs to the technical field of network public opinion prediction, and particularly relates to a network public opinion analysis and prediction method and system based on a space-time diagram neural network. The method comprises the following steps: acquiring network public opinion data, extracting public opinion texts, release time and forwarding, replying and reference relationships, and constructing a bilateral space-time heterogeneous relationship graph containing semantic association edges in time layers and propagation causal edges between the time layers at the same time; further extracting a text representation feature, a network position feature, a propagation statistical feature and a cross-layer propagation feature to form a node input vector, and obtaining a node embedding representation through a space-time diagram neural network; on this basis, a ternary trend state vector composed of propagation activity, propagation attenuation and path offset is constructed, trend classification, propagation direction clustering, key node identification and hot event labeling are completed, and a double-deviation feedback quantity is utilized to carry out iterative correction on an edge weight and a node coding parameter. According to the scheme, structured analysis and trend prediction of the network public opinion propagation process can be realized.

The invention relates to an electric power public opinion data driven work deployment adaptive optimization method and system, and the method comprises the following steps: S1, obtaining multi-source data related to electric power public opinions, filtering noise data, forming a standardized data set, and constructing a unified public opinion data pool; s2, constructing a public opinion knowledge graph, and obtaining a comprehensive public opinion risk index; s3, constructing a multi-target service deployment adaptive optimization model; s4, according to the multi-target service deployment adaptive optimization model, obtaining an optimal work deployment scheme based on a scheme solving and generating method of an adaptive optimization algorithm; and S5, according to the optimal work deployment scheme, pushing the optimal work deployment scheme to a decision maker through a visual billboard, tracking the whole execution process of the scheme, and feeding back an execution result and newly added public opinion data to the step S1. Through real-time public opinion analysis, high-risk events are automatically identified before event expansion, active intervention is realized, and a traditional passive complaint waiting mode is changed.

The invention discloses an information propagation scale prediction method based on multi-feature fusion. The method comprises the following steps: firstly, preprocessing collected information propagation link data and constructing an attention network and a forwarding network; then, extracting multi-dimensional features from three dimensions of a space structure, a time evolution process and a text attribute of information diffusion; and finally, fusing the multi-dimensional features, and inputting the fused multi-dimensional features into a machine learning prediction model for training and prediction so as to realize automatic and accurate prediction of the propagation scale of the information on the social network. According to the method, through multi-feature fusion, the problems of incomplete feature mining and poor prediction precision in the prior art are solved, the information propagation process can be flexibly and effectively quantified, the prediction accuracy and interpretability are remarkably improved, and the method can be widely applied to the fields of network public opinion analysis, hotspot discovery and the like.

The invention discloses a public opinion analysis method and device, equipment and a medium. The method comprises the steps of obtaining historical exchange rate data and historical news data of a target field; based on the historical exchange rate data, screening out target news data from the historical news data; labeling the target news data to obtain labeled data including an emotion label, an importance label and a category label; a plurality of prediction models are trained based on the annotation data, and the prediction models comprise a news classification model, an emotion influence prediction model and an importance degree evaluation model; and analyzing the real-time news data through the plurality of prediction models, and generating a prediction result of the influence of the real-time news data on the transaction on the market. Quantitative analysis of the influence of news events on transactions and markets is realized, so that more accurate and direct data support is provided for transaction decisions.

The application discloses a kind of based on multi-dimensional structure coupling and phase change determination digital content asset evaluation method and system.The method comprises: obtaining multidimensional state data;System state characteristics are constructed;Phase change determination is carried out based on state variable derivative change;When detecting phase change, value transition analysis is carried out;Output value evaluation result.Wherein, the system state characteristics are mapped by multivariate function containing structure coupling degree function, the phase change determination is realized by comparing first derivative or second derivative of state variable with preset threshold, and the value transition analysis is realized by nonlinear model containing piecewise function or exponential function.The application identifies the transition process of digital content from exchange value to structural value, solves the technical problems that the value evaluation dimension is single and the dynamic adaptability is poor in the prior art, and can be widely applied in AI content evaluation, digital asset pricing, public opinion analysis and other fields.

The invention discloses a financial market public opinion analysis method, system and device, a medium and a product, and relates to the field of public opinion analysis, and the method comprises the steps: carrying out the preprocessing of multi-source data, and determining the preprocessed multi-source data; the multi-source data comprises financial market negative public opinion data, city investment data, announcement files, industry and commerce data and enterprise internal business data; based on a large model technology, classifying the preprocessed multi-source data by using a public opinion analysis model special for the financial field, and determining a subject type; the main body types comprise city investment enterprises and non-city investment enterprises; according to the subject type, a structured public opinion analysis report is automatically generated, the method has efficient data processing capacity, the analysis efficiency can be improved, and trend prediction and decision suggestions can be provided.

The invention relates to a time sequence prediction model construction method and system fusing macroscopic factors and public opinion analysis. The method comprises the following steps: acquiring public opinion related data and macroeconomy related data; adopting a vector embedding method to generate a public opinion vector and a macroeconomic vector; analyzing the correlation among the vectors to obtain vector correlation features; obtaining a vector conduction mode through conduction modeling; in combination with a classic economic cycle division standard, a vector cycle conduction mode is obtained through cycle collaborative conduction analysis; and embedding the vector periodic conduction mode into a preset time sequence framework to construct a time sequence prediction model. By adopting the method, the accuracy and robustness of time sequence prediction can be improved by deeply fusing the macroscopic factors and the public opinion dynamics.

The system according to this embodiment aims to understand employee satisfaction in detail and to grasp the state of the organization in real time. [Solution] The system according to the embodiment comprises a question addition unit, an opinion collection unit, an opinion analysis unit, a result aggregation unit, and a solution presentation unit. The question addition unit adds open-ended questions. The opinion collection unit collects opinions from employees. The opinion analysis unit analyzes the opinions collected by the opinion collection unit. The result aggregation unit aggregates the results based on the opinions analyzed by the opinion analysis unit. The solution presentation unit presents solutions based on the results aggregated by the result aggregation unit.

The invention relates to the field of public opinion analysis, in particular to a financial public opinion sentiment analysis method based on multi-modal dynamic weight fusion, and the method comprises the steps: obtaining multi-modal data input in a financial public opinion scene, and carrying out the feature extraction of different modal data input; based on a self-adaptive attention fusion mechanism, dynamically calculating emotional contribution weights of different modes in a current financial scene, and performing weighted fusion on multi-mode input; classifying the multi-modal features by adopting an emotion classifier to obtain emotion polarity and a continuous emotion intensity value; performing abnormal conflict detection and secondary verification on different modal inputs; and performing dynamic confidence coefficient calibration, and writing the emotional intensity value into a time sequence database in real time. According to the financial public opinion sentiment analysis method, the judgment accuracy and robustness of the financial public opinion sentiment analysis method are comprehensively improved through the design of pre-construction of the financial scene template library, the abnormal conflict detection and secondary verification mechanism, the time attenuation factor and the behavior confidence coefficient calibration mechanism.

This application provides an information processing method, apparatus, device, medium, and program product, relating to the field of information technology, for improving the accuracy of identifying opinion leaders. The specific technical solution is as follows: Based on user influence information of a first user within a preset time period, an influence index of the first user is calculated. This user influence information includes user text content data, which includes at least one published text and at least one corresponding comment text from the first user. Based on the at least one published text, a comprehensive sentiment weight is obtained, which characterizes the first user's emotional polarity and intensity. Based on the user text content data, a text resonance degree is obtained. Based on the influence index, comprehensive sentiment weight, and text resonance degree, the first user's opinion leader index is obtained. This application is applied to social media marketing and public opinion analysis scenarios.

The invention relates to the technical field of network public opinion monitoring, and particularly discloses a film and television work network propagation public opinion analysis method, which comprises the steps of multi-modal data acquisition, theme tracking, sentiment analysis and public opinion situation awareness. According to the scheme, a mode of combining LDA topic modeling and hierarchical clustering is adopted, a tree hierarchical structure and a vocabulary evolution relation of a topic are constructed, potential topics of film and television works in network propagation public opinions are learned from word frequency co-occurrence, topic popularity is quantized and mapped to a timeline through dynamic popularity calculation, and therefore the topic popularity is obtained. The interpretability and operability of the result are ensured; the method comprises the following steps: quantifying themes of film and television works by synthesizing popularity fluctuation and emotion dispute, automatically screening key nodes in network propagation public opinions based on a theme tree, quantifying emotion intensity by fusing confidence, and aggregating the emotion intensity to a macroscopic theme and a time window, so as to construct a macroscopic and microscopic integrated dynamic public opinion map; and the sentiment analysis accuracy is greatly improved.

This invention discloses a system and method for calculating and warning about the popularity of online public opinion, belonging to the field of public opinion analysis technology. The method includes the following steps: S1, determining the public opinion object and its corresponding time domain for popularity calculation; S2, constructing basic popularity metrics using the public opinion object and time domain; S3, constructing popularity status metrics using the basic popularity metrics; S4, constructing popularity change description results using the popularity status results; S5, using the popularity change results to make early warning judgments. This invention, by setting up a set of basic popularity metrics, ensures that public opinion popularity calculation is based on a clearly defined public opinion object and time domain. It breaks down and standardizes the composition of public opinion popularity, allowing the scale of public opinion dissemination, the level of interactive activity, and the breadth of participation to be separately characterized and uniformly output, thereby effectively reducing interference caused by differences in data standards between different online platforms and different public opinion events.

The invention relates to the technical field of data processing, in particular to a multi-modal data processing method and system based on dynamic weight fusion, and the method comprises the steps: obtaining multi-modal data, and carrying out the standardization preprocessing; constructing a dynamic weight model based on an attention mechanism, constructing a cross-domain knowledge graph, and performing weighted fusion on the multi-modal features according to a dynamic weight coefficient matrix; establishing a real-time feedback mechanism, and iteratively updating the model by using parameters after feedback optimization; the method has the advantages that the dynamic weight of each modal is calculated in real time through the attention mechanism, and compared with an existing fixed weight method, the data processing efficiency in a complex scene is improved by more than 30%. In a multimedia public opinion analysis scene, the weight can be dynamically adjusted according to text emotion intensity, image visual feature saliency and audio intonation change, and public opinion homes are accurately captured.

The embodiment of the invention discloses a multi-modal public opinion analysis method and system, electronic equipment and a storage medium. The method comprises the following steps: acquiring original multi-modal data associated with a target service; performing data processing on the original multi-modal data based on a data processing mode corresponding to each data type, and determining multi-modal public opinion data corresponding to the target service; performing multi-dimensional public opinion analysis on the multi-modal public opinion data based on a preset public opinion analysis model, determining a public opinion analysis result corresponding to the target business, and associating the original multi-modal data with the public opinion analysis result to obtain a public opinion event corresponding to the target business; and performing decision analysis on the public opinion event based on a preset auxiliary decision mode, determining a disposal decision corresponding to the public opinion event, and pushing the disposal decision to a service system to which the target service belongs. Through the technical scheme of the embodiment of the invention, the richness and diversity of the basic data to be subjected to public opinion analysis and the accuracy and comprehensiveness of public opinion analysis can be improved.

The invention discloses an enterprise ESG public opinion analysis method, electronic equipment and a storage medium. ESG report and annual report content are extracted through keywords and semantics to obtain ESG information, and then public opinions are analyzed. The technical advantages of the scheme are that the method breaks through the limitation of a single text, achieves the collaborative analysis of images, texts and charts, remarkably improves the integrity of extracted information, deeply understands the semantics of the ESG field and the building material industry, achieves the higher correlation matching and theme extraction, fuses the multi-modal emotion information, and enlarges the emotion recognition range.

This invention belongs to the field of natural language processing and deep learning technology, and discloses a stance detection method and related apparatus based on ELECTRA and multi-task learning. The invention first acquires text-target pair data to be detected, and obtains standardized samples through regular expression cleaning, word segmentation, and truncation. The standardized sample format is adjusted to frame input samples, which are then input into the ELECTRA model's pre-trained encoder to extract the ELECTRA model's encoded feature vectors. Local features are extracted, and a multi-task learning framework is constructed through a shared bottom-layer and private upper-layer expert layer structure, encompassing the main task of stance detection, the auxiliary task of sentiment consistency verification, and the auxiliary task of domain classification. The features of each task are fused, and a weighted loss and two-stage training are used to obtain the stance detection model. The data to be detected is processed through the corresponding steps, and the stance result is output. This invention can improve the accuracy, generalization ability, and efficiency of stance detection, meeting the growing demand for online public opinion analysis.

The application provides a multi-modal sarcasm detection method based on hierarchical attention fusion and a graph attention network, belongs to the technical field of social network public opinion analysis and natural language processing, S1: feature extraction and preprocessing, acquires social media image-text data, S2: graph network interaction modeling, constructs a cross-modal heterogeneous graph, S3: sentiment conflict quantification, generates a sentiment conflict score, S4: hierarchical fusion prediction, realizes end-to-end discrimination of sarcasm attributes, and the hierarchical attention fusion mechanism is designed, multi-modal information is fused layer by layer at the feature layer, the relationship layer and the decision layer, different levels of semantic information can be fully utilized, high costs caused by full connection calculation of a traditional complex interaction model can be effectively avoided, and the expansion capability of the model for processing long texts and high-resolution images is improved.

The application discloses an opinion analysis workflow system, comprising: a data analysis function module, which comprises N data analysis modules capable of analyzing opinion data; a workflow establishment module, which selects multiple data analysis modules from the data analysis function module according to opinion analysis requirements, connects them in sequence, establishes a workflow for analyzing a specific task, establishes multiple workflows for analyzing multiple opinion analysis tasks according to opinion analysis requirements of different analysis angles for the same event; a workflow management module, which performs data analysis and calculation on the established workflow and views the calculation results through a visual workflow diagram; and an event management module, which manages multiple opinion analysis tasks for the same event and builds a connection between different tasks through opinion analysis data. The application further provides an opinion analysis workflow method. The system and method can obtain opinion information from different levels and different angles according to opinion analysis requirements.

The application discloses an intelligent public opinion analysis method and device, terminal equipment and a storage medium. The method comprises the following steps: collecting user video comment data; generating a concerned video image based on the collected user video comment data, and generating an intelligent public opinion preliminary analysis result based on the generated concerned video image; generating comment analysis information based on the collected user video comment data and the intelligent public opinion preliminary analysis result; and performing video public opinion analysis based on the intelligent public opinion preliminary analysis result and the comment analysis information. The intelligent public opinion analysis method based on comment collection has the advantages of lower cost, more comprehensive coverage, higher efficiency and higher accuracy compared with traditional methods.

The invention discloses a cross-platform information acquisition and public opinion processing optimization method and system. The method comprises the following steps: collecting network public opinion data from a plurality of independent public opinion monitoring subsystems; performing standardized preprocessing on the collected data from different subsystems to generate standard public opinion data in a unified format; performing public opinion analysis on the standard public opinion data based on a multi-dimensional analysis model, and generating an analysis result containing propagation popularity and risk assessment; and generating a public opinion monitoring report and early warning information according to an analysis result. The system comprises corresponding function modules. According to the method, multi-source data are integrated through a unified framework, deep analysis is performed by using a multi-dimensional model, accurate monitoring, intelligent early warning and resource optimization of whole-network public opinions are realized, and the automation level and decision support capability of public opinion processing are effectively improved.

The application belongs to the field of network public opinion analysis, and particularly relates to a social topic propagation prediction method based on positive and negative guidance topics, which comprises the following steps: obtaining topic data by using an API interface of a social network, extracting message internal factor features and user internal factor features according to user topic data; calculating positive guidance message influence and negative guidance message influence; calculating positive guidance message driving force and negative guidance message driving force by using evolutionary game theory; calculating user independent forwarding probability; constructing propagation dynamics equation according to the user independent forwarding probability; solving the propagation dynamics equation to obtain propagation prediction results of users to different guidance type topics; and the application can more accurately predict the guidance topic propagation trend in the social network.

The invention relates to the technical field of public opinion analysis, in particular to a deep learning-based public opinion data analysis method. The method comprises the following steps: collecting public opinion data of each channel, processing the collected public opinion data, carrying out subject classification on the processed public opinion data, determining text data corresponding to each classified subject, converting the text data subjected to subject classification into a numerical vector, carrying out operation on the numerical vector, extracting deep information from the public opinion data, and extracting the deep information from the public opinion data. Performing statistical summarization on the deep information to generate an analysis result, judging whether the public opinion data analysis meets the standard or not based on the public opinion data analysis duration, judging whether the public opinion data analysis meets the standard or not based on the chain complete index proportion, and adjusting or giving out a corresponding alarm according to corresponding parameters. According to the method and the device, accurate monitoring for the analysis result of the public opinion data is effectively realized, intelligent adjustment for corresponding parameters is effectively realized, and the analysis efficiency of the public opinion data is effectively improved.

This invention discloses a risk-based public opinion analysis method based on natural language processing, comprising: real-time capture of public opinion text; desensitizing user identity fields using hash pseudo-identifiers to obtain the original text stream; constructing a dual-classification weakly supervised framework to calculate risk confidence; when the risk confidence is ≥ a first threshold, writing the corresponding text into a labeling pool to form a risk corpus; mapping the risk corpus into a dynamic heterogeneous risk knowledge graph according to entity-relationship-time slice meta-path rules; applying a multi-head spatiotemporal diffusion Transformer to the subgraphs intersecting with the current window in the graph to obtain the risk propagation vector of the event node; inputting the risk propagation vector and the historical baseline vector into a causal comparison analysis model to output an early risk warning index; when the warning index is ≥ a second threshold, pushing a risk alarm containing the event path to the backend control center. This invention effectively improves the automation, accuracy, and reliability of public opinion risk monitoring and early warning.

The invention discloses an international market public opinion analysis and business opportunity mining system based on deep learning, and relates to the technical field of cross-language emotion calculation, causal reasoning and anomaly detection. The system comprises a multi-source public opinion data acquisition module, a cross-cultural emotion recognition module, a causal graph neural network analysis module, an emotion anti-fact simulation module, a cross-context nested causal chain business opportunity discovery module, an abnormal emotion group detection module, a causal graph generation type learning engine and a business opportunity early warning output module. Through cooperative work of all the modules, the system can automatically identify hidden causal relationship chains from public opinion data of different languages in the world, predict potential market influences in advance, discover commercial opportunities such as supply gaps and sudden demand increase, and give an early warning for abnormal public opinions. According to the system provided by the invention, the depth and accuracy of international market public opinion analysis are remarkably improved, and interpretable business opinion intelligence and decision suggestions can be provided in time.

The invention relates to the technical field of natural language processing and text mining, and discloses a dynamic granularity timeline abstract generation method based on multi-agent collaboration and application thereof.The dynamic granularity timeline abstract generation method comprises the steps that firstly, a high-quality text set related to a query theme is screened out from a multi-source heterogeneous text through cooperative work of three kinds of agents including generation, debate and decision; then dynamically dividing the text into a specified number of semantic clusters through cooperation of another group of agents based on granularity parameters specified by a user; generating an event abstract for each cluster; and finally, carrying out fact consistency verification. According to the method, the accuracy and interpretability of information screening and event grouping are improved through a multi-agent debate mechanism, flexible dynamic control over abstract granularity is achieved by introducing granularity parameters, the content reliability is guaranteed through authenticity verification, the capacity of automatically generating reliable and customizable timeline abstracts from complex text streams is improved, and the method is suitable for popularization and application. The method is suitable for multiple scenes such as public opinion analysis, event traceability and decision support.