228 results about "Stream analytics" patented technology

The invention relates to a multi-modal fusion real-time digital human driving method based on unified behavior vector mapping, and belongs to the technical field of human-computer interaction. The method comprises the following steps of: 1) extracting multi-modal features of voice, action and vision, and generating input features by adopting language feature-emotion decoupling, staged action modeling and macroscopic-micro expression flow analysis; (2) cross-modal collaborative alignment and conflict correction are carried out, and high-precision fusion is realized through knowledge graph constraint time sequence decomposition, particle-by-particle interactive fusion and a situation-driven modal dominant strategy; and 3) constructing a three-dimensional behavior vector space, mapping the multi-modal features to a unified coordinate by utilizing honeycomb grid projection, and driving the digital human to output in combination with a coordinate-action mapping table. According to the method, the problems of difficulty in multi-modal time sequence alignment, feature isomerism and out-of-control conflict are solved, real-time interaction of naturalization, emotional and scenarized is realized, and the expressive force and adaptability of the digital human are remarkably improved.

The invention discloses an intelligent archive opening and identifying method based on a large model, particularly relates to the technical field of archive data auditing, and is used for solving the problems of insufficient cross-modal data analysis capability, lagging rule updating and low man-machine cooperation efficiency in the prior art. Fusing cross-modal features of texts, images and metadata through a hybrid expert model to generate multi-modal feature vectors, and dynamically allocating the multi-modal feature vectors to a rule network, a semantic network and a domain network for cooperative processing based on attention weights; the rule network parameters are optimized through gradient projection constraint, and regulation-driven real-time adaptation is achieved; matching sensitive data in combination with a multi-dimensional feature matrix of auditing personnel, and optimizing task allocation accuracy; removing redundant links by utilizing value flow analysis to generate a lightweight process, and recording as a tamper-proof evidence chain through a block chain evidence storage solidification operation; the auditing efficiency and accuracy are improved, and the compliance traceability is guaranteed.

The invention provides a credit risk dynamic assessment and early warning method based on multi-dimensional data analysis, and relates to the technical field of financial risk control and big data analysis, and the method comprises the steps: obtaining multi-dimensional data from borrower basic information, financial statements, credit records and an external market environment, carrying out cleaning, standardization processing and storage on the multi-dimensional data; based on the multi-dimensional data, through a credit scoring model, cash flow analysis, liability ratio analysis and market environment risk assessment, generating a financial risk level of the borrower; the financial and market data of the borrowers are updated in real time, the dynamic change of the risk state of the borrowers is monitored in combination with a time sequence prediction model and risk correlation analysis, and a dynamic monitoring result is generated; and according to the risk level and the dynamic monitoring result, triggering a grading early warning mechanism, and generating a corresponding control measure suggestion so as to adjust a credit strategy or adopt a risk slow release means.

The invention discloses a space-ground integrated dynamic traffic flow analysis and road topology reasoning method and device. A cloud server generates and issues real-time traffic information through multi-level processing such as map matching, traffic flow analysis, prediction and early warning and road topology analysis and updating; specifically, firstly, the high-precision positioning and communication capability of a low-orbit satellite is utilized, data fusion of an inertial measurement unit and a wheel speed sensor is combined, and a multi-source fusion algorithm is constructed to improve the positioning precision; secondly, traffic flow dynamic analysis and prediction are carried out based on a Transform interaction analysis model, and real-time updating of a road topological structure is realized at a cloud end by using a clustering algorithm and a graph theory model; finally, the traffic flow and road topology information is sent to related vehicles and a traffic management system, and real-time monitoring, early warning and optimization of the traffic state are achieved. According to the invention, the accuracy and real-time performance of traffic information acquisition, processing and issuing can be effectively improved, and the intelligent level of traffic management and decision making is enhanced.

The invention provides an AI-based carbon asset and carbon inspection management method and platform, and belongs to the technical field of carbon emission management, and the method comprises the steps: inputting the operation activity data of a target object into a preset energy efficiency analysis model, and obtaining an energy efficiency analysis result; based on the energy efficiency analysis result, performing energy flow modeling and analysis on the operation activity data of the target object to obtain an energy flow analysis result; the energy flow analysis result is input into a preset energy efficiency balance optimization model, an optimized energy distribution scheme and carbon emission prediction values, predicted based on the energy distribution scheme, of all links are obtained, and the energy efficiency balance optimization model is configured as follows: under the condition that the production constraint condition of the target object is met, the energy distribution scheme is optimized; optimizing energy distribution by taking minimization of the total carbon emission as a target; comprehensively calculating the carbon emission predicted value of each link to obtain a target total carbon emission predicted value of the target object; and generating a corresponding target carbon asset management strategy according to the target total carbon emission prediction value, the carbon transaction market data and the carbon sink data.

The invention discloses a space-time carbon emission prediction method, system and device and a medium. The method comprises the steps of obtaining load historical sequence data; carrying out load flow calculation by adopting a first load flow analysis algorithm to obtain a load flow calculation result; based on the load flow calculation result, carbon flow analysis is carried out through power transmission between the nodes, and the carbon emission coefficient of each node is obtained; constructing a time sequence diagram neural network model, and performing preliminary training on the time sequence diagram neural network model; performing hyper-parameter optimization on the preliminarily trained time sequence diagram neural network model by adopting a first hyper-parameter optimization technology to obtain an optimized time sequence diagram neural network model; and inputting load sequence data and unit power output data at the current moment into the optimized sequence diagram neural network model, and carrying out online prediction on a carbon emission coefficient. According to the method, the real-time prediction of the carbon emission coefficient is realized based on the characteristic learning of the historical operation and monitoring data of the power system.

The invention relates to a house building construction stage carbon emission dynamic estimation method and system based on full life cycle data, and the method comprises the steps: dividing a construction region based on a BIM model, laying a multi-source sensor network, fusing static BIM data and a dynamic Internet of Things monitoring flow, building a four-dimensional space-time label data set, and carrying out the construction stage carbon emission dynamic estimation based on a double-layer dynamic correction model. Kalman filtering is utilized to eliminate instantaneous interference, LSTM transfer learning is combined to predict mechanical efficiency attenuation, a dynamic correction coefficient matrix of a mechanical aging rate, operator skills and environmental factors is embedded, and a carbon emission value is calibrated in real time; when abnormal emission is detected, a key pollution source is traced and positioned through a graph neural network, a three-dimensional thermodynamic diagram is generated in the digital twinborn model, and component-level carbon flow analysis and early warning are performed, so that the carbon emission estimation precision is remarkably improved, the abnormal traceability time consumption is shortened, and the timeliness is remarkably improved.

The invention relates to the technical field of data processing, and discloses a BNCT digital twinborn model construction method, a BNCT digital twinborn model dynamic optimization method, BNCT digital twinborn model dynamic optimization equipment and a medium, and the construction method comprises the following steps: carrying out data assimilation processing on obtained multi-physical field simulation data to generate equivalent measured data matched with physical characteristics of historical measured data, generating a data assimilation training set based on the historical measured data and the equivalent measured data; wherein the multi-physical field simulation data represents simulation data of the accelerator and the beam shaping device in various physical processes; based on the data assimilation training set, a first sub-model used for representing the physical process of the accelerator and a second sub-model used for representing the neutron beam analysis process are obtained through training; and constructing the BNCT digital twinborn model according to the first sub-model and the second sub-model obtained by training. According to the technical scheme provided by the invention, the calculation precision and the processing complexity of the BNCT digital twin system can be improved, and real-time response is ensured.

The multi-source data automatic collection method based on the large language model comprises the steps that feature distribution of historically collected data is analyzed, and an optimal collection strategy is dynamically generated in combination with the state of a current data source; scheduling the acquisition task in real time to obtain a dynamically scheduled data stream; performing real-time anomaly detection and self-adaptive compensation on the dynamically scheduled data stream to obtain a self-fault-tolerant data stream; analyzing entity association and service logic of the self-fault-tolerant data stream, and predicting potential data demand points; automatically expanding a data acquisition range to obtain knowledge-enhanced data assets; identifying protocol features of the data transmission stream of the data assets, and analyzing the data by an analyzer based on protocol feature mapping to obtain protocol-independent standardized data; and performing semantic noise filtering and cross-modal cleaning to obtain semantic pure data. According to the invention, the defect that protocol adaptation and adaptive strategy optimization cannot be carried out in the existing multi-source data acquisition technology is overcome.

The invention relates to the technical field of urban traffic management and carbon emission reduction accounting, and provides a bus travel carbon emission reduction accounting method fused with Beidou spatio-temporal data, which comprises the following steps: acquiring spatio-temporal trajectory data based on a Beidou satellite navigation system, and constructing a passenger-trajectory-station three-dimensional correlation model for passenger flow analysis; constructing a passenger transfer analysis system based on the passenger flow analysis result; constructing a dynamic'perception-prediction-optimization 'dynamic scheduling optimization system for passenger flow detail analysis on the basis of space-time trajectory data and a reinforcement learning algorithm; and constructing a mileage-frequency distribution model and a station efficiency evaluation model based on the spatio-temporal trajectory data, respectively generating a trajectory matching optimization strategy and a differentiated station optimization scheme, and realizing adaptive optimization of the public transportation system under dynamic demand change. Compared with a traditional static data dependence method, the method achieves the dynamic and fine-grained monitoring of the operation process of the public transportation system, and effectively improves the timeliness and integrity of data.

The invention relates to the technical field of micro-grid scheduling, and discloses a time-sharing carbon flow-based micro-grid scheduling method, system and device and a storage medium, and the method comprises the steps: inputting power load data and characteristic data into a carbon flow calculation model, and calculating the carbon flow of a micro-grid; according to the energy consumption of the enterprise, calculating other energy carbon flow of the enterprise, and according to the power consumption period, adding the micro-grid carbon flow and the other energy carbon flow to obtain a time-sharing carbon flow data sequence; and performing carbon flow analysis on the time-sharing carbon flow data sequence, determining a scheduling period, constructing an optimal scheduling model of the micro-grid according to the scheduling period and the carbon flow calculation model, and solving the optimal scheduling model to obtain an optimal scheduling strategy. Optimized scheduling is carried out according to carbon emission conditions of different time periods, organic unification of operation of the micro-grid and operation of an enterprise is achieved under the low-carbon target, and carbon emission of the enterprise is indirectly reduced by improving the participation degree of new energy elements at the specific moment and optimizing the energy structure.

Disclosed in the present invention is a people flow analysis system based on Internet of Things information acquisition. The system comprises a data acquisition module, a data storage module, a people flow statistics module, an obstruction analysis module, a coupling analysis module and a predictive analysis module. The data acquisition module acquires all people flow data of a laboratory server room by means of an image acquisition apparatus; then, the people flow statistics module separately performs crowd density analysis, wherein an average distance is determined by means of hierarchical and clustering analysis of head labeling points, and further a Gaussian function is used to determine the size of heads, thus solving the problem of uncertainty caused by a nearest neighbor method when crowd densities are low and uneven; the obstruction analysis module performs feature fusion on features of an obstacle in a coverage space and neighboring crowds to obtain a new data set; the predictive analysis module performs overall people flow predictive analysis on the basis of the new data set and people flow data of crowds not adjacent to the obstacle. Therefore, the accuracy of people flow analysis and control efficiency are greatly improved.

The invention discloses a deep reinforcement learning optimization method for injection molding process parameters, and belongs to the technical field of intelligent manufacturing. The method comprises the following steps: constructing a dynamic causal graph network through information entropy flow analysis and transfer entropy calculation, and revealing a causal relationship and time delay characteristics among process parameters; manifold learning is adopted to map a high-dimensional parameter space to a low-dimensional manifold, and Riemannian metric guide optimization search is constructed based on the quality gradient; generating enhanced state representation fusing causal association and manifold geometric information; identifying a production element state and selecting a corresponding optimization strategy; a geodesic line is planned in a manifold space to obtain an optimal parameter adjustment path; historical experience is utilized through memory retrieval and case adaptation; cross-task knowledge migration is realized; adopting a depth deterministic strategy gradient algorithm to optimize the decision; and online learning is realized through elastic weight consolidation. According to the method, the problems of black box decision, slow convergence, difficulty in knowledge reuse and the like in the prior art are solved, the optimization efficiency and the interpretability are improved, and the method has the capability of quickly adapting to new tasks.

The invention discloses an intelligent data storage management system which comprises an intelligent data management platform, the intelligent data management platform comprises a pooling storage identification module, a data classification module, a data migration module and a flow-saving analysis module, and the invention relates to the technical field of data management. According to the intelligent data storage management system, hot data, cold data and cloud backup are matched, the data are stored in a proper position according to the access frequency and the use requirement of the data, the data storage efficiency and the access speed are improved, meanwhile, the safety and the reliability of the data are ensured, and through access frequency monitoring, frequency sorting and dynamic standard drawing, the data storage efficiency is improved. According to the method, classification can be automatically carried out according to the access frequency of the data, the accuracy and reliability of the data are ensured while the working efficiency is improved, powerful support is provided for intelligent migration of the data, and intelligent management of data storage is achieved while the method has the advantages of being high in storage efficiency, high in access speed, intelligent in backup and intelligent in migration.

The invention discloses an interrupt-driven program defect detection method based on neural symbol fusion, and relates to the technical field of interrupt-driven program defect detection, which comprises the following steps: step 1, a concurrent analysis stage of supporting interrupt: aiming at a program P to be analyzed, carrying out parallel analysis on the program P; a program point set S with defects is recognized through preliminary static analysis, an information set Q which is difficult to process accurately through static analysis is collected, and a foundation is laid for subsequent analysis; the method specifically comprises the following three sub-steps: sensitive point analysis: positioning a program point set S with defects in a program by utilizing a data flow analysis technology according to a defect type T to be detected; according to the interrupt-driven program defect detection method based on neural symbol fusion, the semantic comprehension ability of a large language model and the comprehensive path coverage ability of static analysis are fully combined, and the limitation of a traditional detection method in the aspects of processing loop logic, array subscript analysis, bit operation solving and the like is effectively overcome.

The invention discloses a blade meridian flow field prediction method based on explicit-implicit mechanism mining coding, and relates to the technical field of turbine blade design, the method comprises the following steps: obtaining meridian plane grids and design working condition parameters of a blade runner on an axial-flow compressor described by using a graph data structure; utilizing the trained blade meridian flow field prediction model to predict physical field data of each node of the meridian plane grid of the blade flow channel according to the meridian plane grid of the blade flow channel and the design working condition parameters; according to the method, the implicit mechanism mining mechanism is used for mining the implicit differential association relationship in the physical data as the data feature in the data feature extraction process of the graph neural network model, and the turbine flow analysis mechanism is explicitly encoded into the model loss function; based on the model loss function, a strategy for optimizing the blade meridian flow field prediction model through back propagation is trained, the blade meridian flow field prediction precision is improved, and the model generalization and the result confidence coefficient are improved.

The invention discloses a cross-province stream pulling analysis method, which comprises the following steps of: determining provinces suspected of active stream pulling and passive stream pulling by analyzing the inflow and outflow of IDC (Internet Data Center) and MAN (Mobile Area Network) of each province; for the province suspected to actively pull the flow, determining an address field suspected to pull the flow by analyzing a downlink flow rate TOPn address field of the province suspected to actively pull the flow; for the suspected pull flow address field, flow velocity and flow proportion analysis is carried out, and a suspected pull flow address pair is determined; carrying out pull flow message feature analysis on the suspected pull flow address pair; and if the pull flow message characteristics are met, determining that the address pair has the pull flow behavior. According to the method, the pull flow behavior is determined according to the settlement rule, the change of the flow and the message characteristics of the pull flow, and the accuracy of cross-province pull flow analysis is improved.

The invention relates to the technical field of intelligent navigation, in particular to a position guide device and method based on mobile cloud passenger flow data, and the device comprises a cloud server which is used for collecting and processing anonymous passenger flow data from a mobile terminal in real time and generating a dynamic thermodynamic diagram; the indoor and outdoor positioning module determines the current position of the user through a multi-source fusion positioning technology; the navigation terminal is used for generating an optimal navigation route through a path planning algorithm according to the user preference and the real-time passenger flow density and outputting navigation guidance; according to the method, the real-time passenger flow analysis and dynamic path planning technology is adopted, the congestion area can be accurately predicted and avoided, the navigation efficiency is greatly improved, and special requirements such as barrier-free navigation are supported by combining an AR (Augmented Reality) navigation interface with a personalized recommendation algorithm, so that different user groups can obtain customized navigation experience.

A system and method for privacy-preserving workflow analysis that monitors user interactions with an application interface while maintaining data privacy. The system generates event data that excludes sensitive content while preserving workflow sequence information, enabling reconstruction and analysis of user workflows without exposing protected information. Event data is stored and used to reconstruct workflow sequences, which can be presented through a user interface for analysis, training, and compliance monitoring. The system supports multiple privacy contexts and can substitute synthetic data during workflow reconstruction, allowing organizations to analyze business processes without compromising sensitive information. The system enables comprehensive workflow monitoring and analysis while addressing privacy requirements in regulated environments, supporting use cases including quality assurance, training, and process optimization.

A method for providing automated agent performance improvement includes obtaining a ranking for each agent of a plurality of agents; obtaining customer-agent interactions for each agent; determining a task label for each of the customer-agent interactions; selecting a first set of customer-agent interactions from the customer-agent interactions corresponding to a first task label; receiving application event streams corresponding to the first set of customer-agent interactions and comprising agent application usage data; analyzing the application event streams for statistically relevant differences for one or more activities performed by a first group of agents compared to a second group of agents; determining at least one activity of the one or more activities that is performed by the first group of agents differently than the second group of agents; generating guidance corresponding to the least one activity for the second group of agents; and deploying the guidance to the second group of agents.

The invention relates to the technical field of rail transit passenger flow analysis, in particular to a rail transit short-time passenger flow prediction method and device based on multi-scale convolution and a storage medium. Forming a multi-source data set through historical time sequence passenger flow data, station network topology data, external environment data and the like, and extracting fusion feature data; extracting multi-scale time characteristics of the passenger flow by using a multi-scale convolution module, and reserving bottom information in combination with residual connection; the over-smoothing problem of the deep network is relieved through the improved spatial dependency relationship between the residual image convolutional network modeling sites; and dynamically fusing the spatial-temporal features and the external environment features by adopting an attention mechanism, and generating a prediction result through a long-short term memory network. According to the method, the passenger flow prediction precision and the model robustness are remarkably improved, the method can adapt to complex external environment changes, and high-reliability decision support is provided for rail transit dispatching management.

Systems, computer program products, and methods are described herein for data protection in a distributed network via reconstruction and analysis of data segments. The present invention is configured to initiate a network flow analysis to assign indicators to specialized access nodes of a network diagram, execute a first cut to remove predetermined nodes, store the corresponding discrete data segments for each of the first plurality of electronic communications at the nodes, construct a first arrangement of a selected group of the corresponding discrete data segments, determine a presence of sensitive information, and construct additional arrangements of the selected group of the corresponding discrete data segments to determine a presence of sensitive information.

The invention relates to the technical field of public transport organization, in particular to an urban traffic station connection evaluation method based on multi-source data, which comprises the following steps of: performing passenger flow analysis on an upstream public transport section of a station, dividing passenger flow time periods, and acquiring multi-source connection situation data by adopting a sampling or real-time monitoring mode based on different time periods; calculating a transport capacity supply saturation index based on the station upstream public traffic section passenger flow data, calculating a dynamic transport capacity service efficiency index based on the mobile passenger-carrying end parking characteristic data, weighting to obtain a comprehensive connection bearing coefficient, estimating the real-time passenger flow arrival flow based on the arrival shift data of the traffic station and the number of resold tickets, and determining the real-time passenger flow arrival flow. And evaluating the urban traffic connection capability of each station by generating a real-time connection efficiency value. According to the method, the quantitative evaluation of the operation state of the connection system of the urban traffic station is realized, the defects of a traditional method in the aspect of multi-source data fusion are overcome, and the analysis precision of the connection capability of the urban traffic is improved.

A method, apparatus, and system are disclosed. The method includes constructing a static single assignment (SSA) form and a static single use (SSU) form for a program; setting a single-use disqualifying property locally for each SSU version, and propagating the single-use disqualifying property both forward and backward on an SSU graph uniquely formed from the constructed SSU form so it becomes a global property; transferring results from the SSU form to the SSA form to set a single-use property locally for each SSA version based on an occurrence of any use being associated with a disqualifying SSU version; performing data flow analysis on an SSA graph uniquely formed from the constructed SSA form so the single-use property becomes a global property to identify one or more definitions as dynamic single-use for variables in the program; and generating computer-readable instructions for executing the program based on the one or more definitions identified as dynamic single use for the variables in the program, wherein the one or more definitions identified as dynamic single-use has a defined value used exactly one time during execution of the program.

The application provides a multi-source data automatic acquisition method based on a large language model, which comprises the following steps: analyzing the feature distribution of historical acquisition data, combining the state of the current data source, and dynamically generating an optimal acquisition strategy; scheduling the acquisition task in real time to obtain a dynamically scheduled data stream; performing real-time anomaly detection and adaptive compensation on the dynamically scheduled data stream to obtain a self-fault-tolerant data stream; analyzing the entity association and business logic of the self-fault-tolerant data stream, predicting potential data demand points; automatically expanding the data acquisition range to obtain knowledge-enhanced data assets; identifying the protocol characteristics of the data transmission stream of the data assets, parsing the data based on a protocol characteristic mapping parser to obtain protocol-independent standardized data; performing semantic noise filtering and cross-modal cleaning to obtain semantic pure data. In the application, the defects that the existing multi-source data acquisition technology cannot perform protocol adaptation and adaptive strategy optimization are overcome.

Systems and methods for performing streaming analysis on a server-less computing system are described. A user may specify a data stream for which analysis should be performed, a server-less function for analysis, and criteria for analysis, such as windowing criteria and aggregation criteria. The windowing criteria may specify windows of items within the stream that should be co-analyzed. The aggregation criteria may specify how items within each window are grouped during analysis. A poller device may read data items from the stream, windowing and grouping the items according to windowing criteria and aggregation criteria, and invoking a server-less function to stream the data items. The poller device may also maintain a state between calls.

The invention provides a college affair assistant construction method and system based on an agent technology. The method comprises dynamic memory flow analysis: (1) an agent memory flow mechanism; (2) memory module interaction; and (3) carrying out personalized analysis. (4) carrying out multi-source information fusion analysis; and (5) multi-agent collaboration and retrieval enhancement. (6) performing multi-agent collaboration; (6) constructing a multi-agent collaboration workflow; (7) a dynamic memory flow energizing agent; and (8) an intelligent agent fusion mechanism. (9) predicting a dual-stage load condition; (10) carrying out progressive load recovery; and (11) a decision is recovered in combination with a multi-dimensional load. The system comprises a personalized analysis module based on dynamic memory flow and student information, an RAG retrieval strategy module based on student information enhancement, an agent design strategy module and a dynamic load recovery mechanism module. The invention provides a high-performance efficient transaction assistant which meets individual requirements for users.

The invention discloses a medical image recognition method and system based on artificial intelligence, and the method comprises the steps: obtaining CT image data, pathological section data and PET metabolic activity data, introducing a GNN graph neural network to model the semantic association between multi-modal features, and outputting a fusion feature vector; inputting the fusion feature vector into a DSRN double-flow space-time recursive network, extracting 3D lesion morphological features of a single image by using a spatial flow, analyzing lesion growth kinetic parameters of a historical image sequence through a time flow, and generating a lesion malignancy probability index based on a gating fusion unit; based on the focus malignancy probability index, Monte Carlo Dropout sampling is utilized to generate a confidence interval, a clinically interpretable credibility score is output, and when the confidence is smaller than a threshold value, a low-confidence area is displayed for a doctor to check. The deep fusion of multi-source information is realized, the one-sidedness of a single mode is avoided, and the recognition accuracy and efficiency are improved.

The invention discloses a meeting place people flow analysis, statistics and prediction system, and relates to the technical field of intelligent people flow management, and the system comprises a face recognition verification terminal, a people flow statistics terminal, a prediction control terminal, an early warning prompt module, an emergency control module, a data storage module and a visualization module. Statistical errors caused by repeated entering and exiting are avoided, the number of people in the museum is calibrated in real time, the problems of identity confusion and insufficient precision existing in traditional manual statistics and single sensor counting are solved, the accuracy and reliability of people flow statistics are improved, and a high-quality data basis is provided for subsequent prediction and control. Historical activity data, real-time reservation information and environment variable factors are fused, a prediction model is constructed to pre-judge the peak period and the peak scale of people flow in advance, the management and control lagging risk caused by a traditional afterward response mechanism is reduced, and intelligent conversion from passive interception to active dredging is achieved in combination with a gate flow limiting and shunting guide strategy.

Analytics equipment (14) for a communication network (10) stores a stream (18) of data records (20) from the communication network (10). The analytics equipment (14) samples the stored stream (18T) to obtain a sampled stream (18S) that includes fewer data records (20) than the stored stream (18T). The analytics equipment (14) explores the sampled stream (18S) to identify characteristics (28) of data records (20) to be used for insight creation. Based on the identified characteristics (28), the analytics equipment (14) filters the stored stream (18T) to obtain a filtered stream (18E) that includes data records (20) with the identified characteristics (28). The analytics equipment (14) then creates one or more insights (16) about the communication network (10) using the filtered stream (18E).