39 results about "Automatic inference" patented technology

Certain embodiments of the present invention provide systems and method for automatic inference and adaptation of a virtualized computer environment. Certain embodiments of a system include a virtual topology and traffic inference framework tool adapted to monitor traffic for an application on a virtual network to produce a view of network demands for the application. The system also includes a monitoring tool adapted to monitor performance of an underlying physical network associated with the virtual network using traffic for the application. Further, the system includes an adaptation component adapted to automatically adapt the application to the virtual network based on the measured application traffic, the monitored network performance, and one or more adaptation control algorithms.

A “similarity quantifier” automatically infers similarity between media objects which have no inherent measure of distance between them. For example, a human listener can easily determine that a song like Solsbury Hill by Peter Gabriel is more similar to Everybody Hurts by R.E.M. than it is to Highway to Hell by AC / DC. However, automatic determination of this similarity is typically a more difficult problem. This problem is addressed by using a combination of techniques for inferring similarities between media objects thereby facilitating media object filing, retrieval, classification, playlist construction, etc. Specifically, a combination of audio fingerprinting and repeat object detection is used for gathering statistics on broadcast media streams. These statistics include each media objects identity and positions within the media stream. Similarities between media objects are then inferred based on the observation that objects appearing closer together in an authored stream are more likely to be similar.

A system, system, and program for automatically inferring and updating an availability status of a user are provided. At least one current activity of a user is detected, where the current activity includes at least one system level event within a computer environment accessible to the user. A current availability status for a particular communication request is inferred from the current activity of the user in view of multiple availability rules, where the multiple availability rules map current activity to multiple possible availability statuses. In addition, inference of the current availability status is determined based on the identity of the requester of communication with the user, where the multiple availability rules map the requester identity in combination with the current activity to multiple availability rules. Output of the current availability status is controlled, such that the requester of the particular communication request is notified of an availability of the user to respond to the particular communication request.

A system, system, and program for automatically inferring and updating an availability status of a user are provided. At least one current activity of a user is detected, where the current activity includes at least one system level event within a computer environment accessible to the user. A current availability status for a particular communication request is inferred from the current activity of the user in view of multiple availability rules, where the multiple availability rules map current activity to multiple possible availability statuses. In addition, inference of the current availability status is determined based on the identity of the requester of communication with the user, where the multiple availability rules map the requester identity in combination with the current activity to multiple availability rules. Output of the current availability status is controlled, such that the requester of the particular communication request is notified of an availability of the user to respond to the particular communication request.

A method and related system to, among other things, automatically infer answers to all of the ADL questions and the first four questions of the IADL in the home. The inference methods detect the relevant activities unobtrusively, continuously, accurately, objectively, quantifiably and without relying on the patient's own memory (which may be fading due to aging or an existing health condition, such as Traumatic Brain Injury (TBI)) or on a caregiver's subjective report. The methods rely on the judicious placement of a number of sensors in the subject's place of residence, including motion detection sensors in every room, the decomposition of each relevant activity into the sub-tasks involved, identification of additional sensors required to detect the relevant sub-tasks and spatial-temporal conditions between the signals of sensors to formulate the rules that will detect the occurrence of the specific activities of interest.

Methods and system for automated inference of physico-chemical interaction knowledge from databases of term co-occurrence data. The co-occurrence data includes co-occurrences between chemical or biological molecules or co-occurrences between chemical or biological molecules and biological processes. Likelihood statistics are determined and applied to decide if co-occurrence data reflecting physico-chemical interactions is non-trivial. A next node or an unknown target representing chemical or biological molecules in a biological pathway is selected based on co-occurrence values. The method and system may be used to further facilitate a user's understanding of biological functions, such as cell functions, to design experiments more intelligently and to analyze experimental results more thoroughly. Specifically, the present invention may help drug discovery scientists select better targets for pharmaceutical intervention in the hope of curing diseases. The method and system may also help facilitate the abstraction of knowledge from information for biological experimental data and provide new bioinformatic techniques.

The invention discloses a machine learning method for realizing ship anthropomorphic intelligent collision prevention decision. An analog source and an example source are generated by off-line artificial learning, and a collision prevention model for on-line acquiring new collision avoidance knowledge, and a database for storing ship parameters are constructed, and an automatic reasoning mechanism, a calculation unit and an evaluation system are designed. The collision prevention model and the automatic reasoning mechanism are used, knowledge discovery and approximate reinforcement learning strategies are realized through online machine learning, and new collision prevention knowledge is acquired, and a dynamic collision avoidance knowledge base is constructed. An inference engine is usedto invoke the ship parameters and the PIDVCA algorithm of the database through the automatic inference mechanism to realize the intelligent collision prevention decision of the machine. The machine iscapable of acquiring the information and the formalized collision prevention domain knowledge on site through the guidance of the automatic reasoning mechanism, is used to learn and solve new knowledge of collision prevention problems of any meeting scene, and has a perception target and a cognitive target to further formulate a scientific and reasonable collision prevention decision scheme, andfinally has a thinking mode for simulating and surpassing the human to solve complex collision prevention problems.

The invention provides a protocol state machine automatic inference method based on state fusion. The method comprises the following steps of message format extraction, message classification, session abstraction and original state machine construction and the state fusion based on output messages. An extended prefix tree transducer EPTT is adopted in the protocol state machine automatic inference method to describe the session process of a protocol entity, the output messages of a protocol are focused on, the same states in a state machine are fused, testability interaction is carried out on the protocol entity to verify the feasibility of protocol state fusion, the automation of the inference of the protocol state machine is guaranteed, and the accuracy of an inference result is improved.

The invention belongs to the technical field of mobile behavior prediction, and specifically relates to a mobile user location prediction method and device for automatically inferring a social relation. The method comprises the following steps: acquiring individual behavior records of a user from a user mobile behavior log database; inferring a social relation type among users according to the individual behavior records; constructing a user social relation network with the user as a node and the social relation type between two users as a connecting side; utilizing the individual behavior records of the user to construct a discrete moving track sequence of the user in time sequence; generating social relation sub-graphs by utilizing a Jaccard coefficient, constructing a null model, comparing a size relationship of statistical index values of various social relation sub-graphs under a real network and the null model, and determining a user group social relation module; performing userindividual social relation model verification; and respectively establishing a Markov predictor, an acquaintances predictor, a familiar stranger predictor and an output regulator used for predicting the future position of the user. The mobile user location prediction method for automatically inferring the social relation provided by the invention can improve the accuracy of location prediction andprotect the individual privacy of the user.

An enhanced multidimensional data management system significantly reduces the resources required to convert a spreadsheet into an enterprise collaborative system. A spreadsheet may be imported as is without modifications. Additional information that cannot be automatically inferred about the data model used in the spreadsheet is obtained. Dimensions, timescale, and measure data are clearly defined and located. The spreadsheet is kept in the system in the form of both a spreadsheet and as a multidimensional data structure such as an online analytical processing (OLAP) cube or structure. This duality allows the spreadsheet to be manipulated as if it was an OLAP cube, modifying iterations within the dimensions, and the OLAP cube to be computed as if it was a spreadsheet.

The invention discloses a log analysis method and device, which comprises the following steps: collecting an existing log and extracting log information from the existing log; Digitizing log information to generate log data; The log data is reconstructed into a data structure which can be processed by machine learning model. The data structure includes training set and test set. Training and subsequently testing a plurality of machine learning models using a training set and a test set, respectively; According to the test results, the machine learning model is selected. And analyze the newly generated logs using the selected machine learning model. The invention can extract and analyze log information from different log files or different types of log files, so as to automatically infer log contents, free labor force and reduce log analysis cost.

The invention discloses an accusation classification method for dealing with unbalanced classification, which comprises the following steps: acquiring and collecting corpus, preprocessing the case description corpus to obtain the case description corpus related to the accusation; The case description corpus related to a crime is taken as the positive example corpus, and the case description corpusirrelevant to a crime is taken as the negative example corpus, and the training corpus and the test corpus are divided. Under-sampling algorithm is used to extract a number of subsets from the negative instance corpus independently and randomly, and each subset and the positive instance sample are combined into a training corpus subset. Several LSTM-based base classifiers are trained by using several subsets of training corpus. Combined with the classification results of the base classifier, the new crime description is classified. The invention can train a base classifier with high classification accuracy rate under the condition that the number of positive samples is small and the number of negative samples is large, solves the situation that the classification error rate is high underthe condition of category imbalance, and realizes the automatic inference of charges of case description.

The embodiment of the invention discloses an inference model training method and device, an electronic device and a storage medium, and the inference model training method comprises the steps: obtaining the current question state of a current question by aiming at the current question in a training sample; selecting a current to-be-executed action according to the current question state; executingthe current to-be-executed action to obtain a new topic state; taking the new topic state as the current topic state of the current topic; according to the technical scheme, the inference model can continuously accumulate the problem solving experience, the inference accuracy is optimized, and the trained inference model can achieve automatic inference of complex elementary mathematical problemsin actual use.

The invention relates to organic projects. A collaborative project management system is provided to create a project workspace linking together various work-product items including e-mails, documents and tasks based on the identification and association of related information contained in such items. A new project workspace may contain groups of documents, correspondence such as emails, instant message and web chatting, tasks, and other data types linked together by common content topics. The system may initiate a new project workspace based on information contained in an initial correspondence, such as an email conversation, by identifying items that are related to the email conversation such as tasks or documents that a user creates or shares subsequent to receipt of an email or message, and automatically inferring an association based on their contents.

The invention discloses a log analysis method and device, comprising: collecting existing logs and extracting log information from the existing logs; Digitizing log information to obtain log data; Thelog data is reconstructed into a data set which can be processed by machine learning model. The data set includes training set and test set. Training a plurality of machine learning models using a training set; Test set is used to test the machine learning model and select according to the test results. Use the selected machine learning model to analyze the new log. The invention can extract andanalyze log information from different log files or different types of log files, so as to automatically infer log contents, free labor force and reduce log analysis cost.

The invention provides a software quality demand modeling and inference method which is used for providing software quality demand oriented formalized modeling, automatic inference and an analysis platform r sigma. The method includes the following steps: setting a related concept, performing graphical modeling, clicking an r sigma drawing area, selecting a corresponding modeling element, clicking a canvas to generate the corresponding modeling element, performing formalized modeling, designing according to semantics and grammar of a target modeling language sigma orienting to software quality demand, performing strategy assessment to determine whether an implementation strategy is an acceptable strategy or not, performing strategy exploration to check all possible combinations of design options and classify strategies according to satisfied states of root nodes, and performing model extension to capture incidence relations among different modeling results. By the method, quality demand modeling results in a text format can be generated; no man-machine interaction is needed; in addition, multi-objective decisions of a quality demand implementation strategy are supported, and mutual dependent relationships among models can be captured.

The invention relates to a rule inference method and system for realizing unmanned vehicle navigation obstacle avoidance, and the method comprises the steps: 1, constructing a knowledge rule library,employing a hierarchical inference strategy, defining and constructing a basic rule and a model library, taking the data collected by an unmanned vehicle in real time as the input data of an inferenceengine, and carrying out the inference decision through an automatic inference mechanism, completing the updating of the knowledge rule base; and 2, enabling the inference engine to allow the unmanned vehicle to complete observation, judgment, decision and action processes through inference engine resource configuration, calling of a calculation unit and information interaction with the knowledgerule base and by adopting an inference engine operation mode based on a hierarchical inference strategy, and has autonomous task capability. According to the invention, new knowledge of a task scenecan be learned, so that the unmanned vehicle has autonomous path planning, navigation obstacle avoidance and target identification capabilities, an intelligent reasoning decision can be made scientifically and efficiently, finally, the unmanned vehicle has a decision capability of simulating and surpassing a person to solve a complex problem, and a new method is provided for realizing unmanned equipment intelligence.

The invention relates to a non-cooperative inference method for the physical topology of an Ad Hoc network, and belongs to the technical field of wireless networks. The method comprises the steps of S1, differentiating and positioning nodes in the Ad Hoc network by adopting radio positioning; S2, obtaining the size of a network deployment area according to the node differentiating and positioninginformation; S3, calculating the important communication distance of the nodes according to the size of the network deployment area and the node differentiating and positioning information; and S4, inferring the physical topology of the network according to the node differentiating and positioning information and the important communication distance of the nodes. The method realizes automatic inference of the network physical topology without knowing the relevant parameters such as the node communication power. Compared with the existing methods, the method improves the structural validity ofthe network physical topology and reduces the calculation complexity of the implementation process.

A computer program (10) analyses a specification for a process, by analysing automatically the specification of the process to identify a set of monitoring points, and determine automatically a cost of the monitoring. Measuring the cost can enable the cost to be reduced or the monitoring to be optimised for a given cost. The analysis can include identifying database accesses, and determine automatically how to reduce the amount of database accesses to improve database performance. It can also analyse automatically the specification to identify database accesses which involve database locking, and infer automatically how to reduce the scope and range of the locking, to improve database performance.

The invention provides an ontology-driven workshop unsafe state semantic reasoning method in a digital twin environment. The method comprises the following steps: firstly, establishing a semantic ontology model of an unsafe state of a workshop production site through a network ontology language (OWL), and establishing inference rules of different types of workshop unsafe states by using a semantic web rule language (SWRL); then, encoding the established semantic ontology of the unsafe state of the workshop production site by using an ontology editor Protege; secondly, simulating the unsafe state virtual scene of the twin workshop in Unity 3D, and recording a high-fidelity simulation animation into a video as a data set source of subsequent target detection; then, identifying instances in an unsafe state of a workshop production site by using an instance segmentation algorithm, and mapping the instances into instances of the established ontology; and finally, executing an inference rule by using an inference engine to realize automatic inference of the potential danger state and related information of the workshop production site.

The invention provides an intrusion detection method of a radio frequency identification system based on ontology. Based on an ontology semantic definition rule, a radio frequency identification intrusion detection system and a method of a two-grade defense system are established with the combination of the tracking of radio frequency identification tags and the positioning technology. To accomplish the design of the method, the invention specifically brings forward an ontology-based inference engine to provide the inference capability of tag clone for the radio frequency identification intrusion detection system. By employing the tracking of the radio frequency identification tags and the positioning technology, a model with the combination of semantic position and physical position is designed to fight against tag clone. An advanced engine based on the ontology semantic rule can provide an advanced automatic inference capability for a radio frequency identification misuse detection system (RIDS). When a tag object is duplicated, or a clone attack occurs, the engine can accurately provide the formalized model for intrusion detection.

In accordance with an embodiment, described herein is a system and method for automatic inference of a cube schema from a tabular data for use in a multidimensional database environment. A cube schemainference component can successively perform column-wise splits to calculate information gain ratios between each pair of a plurality of columns in the tabular data. A cross correlation matrix can beconstructed from the information gain ratios between each pair of columns. The system can determine relationships among the plurality of columns based on the cross correlation matrix; create a hierarchy directed graph to represent the relationships; and further transform the hierarchy directed graph into a cube schema that can be used to create a cube for loading the tabular data, or to map the tabular data into an existing cube in the multidimensional database environment.

The invention discloses a tank field oil transportation operation scheduling optimization method and belongs to the computer auxiliary manufacturing field. The method comprises steps that a, the information of a crude oil tank field and the product oil tank field equipment is collected, and basic tank, valve, pump and pipeline attribute configuration is carried out; b, the production data is written into a remote terminal unit, the data is read by a data acquisition server from the remote terminal unit, and the data is written into a database; c, the real-time production data is read, each tank, valve, pump and pipeline use state is determined, and the information is written into the database; d, a connection relationship of the tank field equipment is taken as a base, in combination with an available state of each equipment and constraint conditions, multiple optimized operation lines are acquired through automatic inference; and e, a proper scheduling operation line is selected by scheduling staff, the selected scheduling operation line is submitted to management staff for check, after an operation flow is checked by the management staff, the operation flow is carried out online. The method is advantaged in that a tank field oil transportation operation scheduling instruction can be optimized, oil transportation operation scheduling errors can be further reduced, and work efficiency is improved.