58 results about "Ground fact" patented technology

A computer-implemented method of gathering large quantities of training data from case law documents (especially suitable for use as input to a learning algorithm that is used in a subsequent process of recognizing and distinguishing fact passages and discussion passages in additional case law documents) has steps of: partitioning text in the documents by headings in the documents, comparing the headings in the documents to fact headings in a fact heading list and to discussion headings in a discussion heading list, filtering from the documents the headings and text that is associated with the headings, and storing (on persistent storage in a manner adapted for input into the learning algorithm) fact training data and discussion training data that are based on the filtered headings and the associated text. Another method (of extracting features that are independent of specific machine learning algorithms needed to accurately classify case law text passages as fact passages or as discussion passages) has steps of: determining a relative position of the text passages in an opinion segment in the case law text, parsing the text passages into text chunks, comparing the text chunks to predetermined feature entities for possible matched feature entities, and associating the relative position and matched feature entities with the text passages for use by one of the learning algorithms. Corresponding apparatus and computer-readable memories are also provided.

A business decision modeling system graphically models business decisions using a rule family structure that adheres to Decision Model principles. In tabular representations, for example, two-dimensional rule families have a heading and a body. The heading includes one or more condition fact types and one conclusion fact type. The body includes unique logic statements corresponding with the heading's fact types, with operators and operands indicating how the logic operates on conditions to determine conclusions. All tested conditions fact types must be true for the conclusion fact type to be true. No two rule families can have the same conclusion fact type, and the conclusion fact type of a first rule family can be a condition fact type of a second rule family. Logic statements are atomic and cannot be further decomposed. The semantic and business integrity of the business rules ensure high quality logic traceable to business objectives.

The present invention provides a way to more effectively estimate energy usage by devices such as servers, including traditional mainframes as well as other enterprise systems. One core idea surrounds the fact that when detailed power consumption information is unavailable for a specific configuration, power consumption is computed / projected / estimated by de-rating the maximum power rating by a power adjustment factor that would more closely project the true power consumption. Under this approach, the power consumption is then provided as a range specifying probable expected lower and upper limits for a given device. This range is more meaningful for planning and decision-making and shows the client a more likely best and worst case energy consumption scenario for their specific device(s), and hence is helpful for planning or server comparison decisions. Cooling, energy costs, energy efficiency, and energy expanded per MIPS are also projected and given as a range instead of a single number. This gives the IT decision maker more accurate information upon which to base IT decision-making.

A method and a device for automatic triggering of deceleration of a vehicle for preventing a collision, which is accomplished by the fact that a variable, which represents a probability of collision with another vehicle, must exceed a predefinable threshold value, and the threshold value is variable as a function of the driver's response, the current driving situation, or the ambient situation. To determine the change in the threshold value, signals from a steering angle sensor, a brake pedal sensor, an accelerator pedal sensor, a device for determining the speed of the vehicle, a device for determining the uphill or downhill slope of the road, a device for determining the vehicle yaw rate, a device for determining the vehicle float angle or a device for detecting stationary and moving objects in the vicinity of the vehicle, in particular in the area in front of the vehicle, are analyzed.

A device is disclosed for detection of surface properties or conditions, in particular detection of water, snow and ice and in particular to read surfaces by means of detector means mounted on individual vehicles, and transmitting the data from the vehicle, preferably together with position data of the vehicle, to be used by drivers of other vehicles for warning of slippery road conditions ahead of the vehicle. According to a particular aspect, it has been realised that the fact that polarized light which is mirror reflected by e.g. ice or water on a surface, preserves its polarization, whereas polarized light which is diffuse reflected largely becomes depolarized, may be utilised to separate the two types of reflection and thus provide a much more compact sensor device for surface properties, in which the light may be emitted close to or parallel to the road surface normal.

Robust crowd counting is an important yet challenging task of interest in a number of applications and is compounded by the fact that crowd behavior is unpredictable. Traditional approaches that tackle these issues are mainly classified into two categories: video based recognition and non-image based localization. However, both suffer inherent drawbacks such as being limited to line-of-sight or requiring people to carry specific electronic devices which limit them even within relatively well controlled environments without considering public locations or rapidly emerging events. Accordingly, it would be beneficial to provide a system that implements a device free crowd counting methodology with no impact on network capacity and / or performance.

The invention provides a method for predicting epidemic situation by a spatial heterogeneity-based infectious disease propagation model. If subareas comprising an area 1 and an area 2 exist in a city, a common rule that crowd flow causes the propagation of infectious diseases and the statistical data (obtained in the form of questionnaire and the like) about the dynamic flow of the crowd in a day can be explained according to a formula, and a method with better operability is established on the basis. The invention aims at reasonably describing a fact, namely the flow that people go out early and come home late for work, shopping and the like does not cause the migration of population macroscopically, but can cause transregional diffusion of the infectious diseases. As the daily periodic flow of the population increases the number of infected persons in a net input subarea, diseases are diffused in space to change the development speed of the epidemic situation. Particularly, at the initial stage of the epidemic situation, the transregional diffusion has an extremely high speed. The thought has heuristic significance for searching regional economic problems such as transregional consumption, labor input and export, and the like.

Disclosed herein is a learning system that uses the concept of tossing a token (dart) at a specifically designed game board for individual or team players. Included in this invention are variations of the learning system involving religious, phonetic and phonological awareness, vocabulary, geography, mathematical, historical, and other facts. This system incorporates the concepts of the general game playing skills and tactics of a commonly known game, darts, but further incorporates the concept of using various facts and a point earning scheme to constitute a learning system that while educational is also entertaining. This learning system includes but is not limited to darts and dart boards because the game board can be made from various materials that will allow it to be safely used by children and adults of all ages both indoors and outdoors.

A device is disclosed for detection of surface properties or conditions, in particular detection of water, snow and ice and in particular to read surfaces by means of detector means mounted on individual vehicles, and transmitting the data from the vehicle, preferably together with position data of the vehicle, to be used by drivers of other vehicles for warning of slippery road conditions ahead of the vehicle. According to a particular aspect, it has been realized that the fact that polarized light which is mirror reflected by e.g. ice or water on a surface, preserves its polarization, whereas polarized light which is diffuse reflected largely becomes depolarized, may be utilized to separate the two types of reflection and thus provide a much more compact sensor device for surface properties, in which the light may be emitted close to or parallel to the road surface normal.

The invention relates to the field of intelligent law, in particular to a criminal case criminal name prediction method based on a sequence enhanced capsule network. The method comprises the followingsteps: S1, constructing a training data set, and obtaining fact description of a case and a criminal name penalty result as training data; S2, constructing a sequence enhanced capsule network model and carrying out training through training data; and S3, through the sequence enhanced capsule network model trained in the step S2, inputting the fact description text of the new case into the sequence enhanced capsule network model, and enabling the model to automatically predict a corresponding criminal name as a criminal name prediction result. The model provided by the invention not only can better capture the significant characteristics and semantic information of legal texts, but also has better competitiveness on the low-frequency criminal name prediction problem; and a final loss function is introduced to serve as a loss function of the sequence enhanced capsule network model, so that the problem of high imbalance of the criminal names of the low-frequency criminal name predictiontask is further relieved.

Automatic relationship extraction is provided. A machine learning approach using statistical entity-type prediction and relationship predication models built from large unlabeled datasets is interactively combined with minimal human intervention and a light pattern-based approach to extract relationships from unstructured, semi-structured, and structured documents. Training data is collected from a collection of unlabeled documents by matching ground truths for a known entity from existing fact databases with text in the documents describing the known entity and corresponding models are built for one or more relationship types. For a modeled relationship-type, text chunks of interest are found in a document. A machine learning classifier predicts the probability that one of the text chunks is the entity being sought. The combined machine learning and light pattern-based approach provides both improved recall and high precision through filtering and allows constraining and normalization of the extracted relationships.

The invention provides an automatic small insect trapping detection method and system thereof. The method comprises the steps that target insects in a collected image are identified and marked; a local minimum algorithm is used to separate partially overlapped target insects in the marked collected image to acquire a first separation and counting result; a local region approximate estimation algorithm is used to separate seriously overlapped and covered target insects in the marked collected image to acquire a second separation and counting result; and a total counting result which comprises the first separation and counting result and the second separation and counting result is acquired. According to the invention, the counting accuracy of the target small insects can be remarkably improved; the target insects which are partially overlapped, seriously overlapped or covered by barriers can be separated to acquire the counting result closest to facts; the environment adaptability and the collected image quality of small injurious insects are improved through an image collecting device.

A method and apparatus that provides an accurate estimate of the time and amplitude of arrival of the first arriving overlapping multipath components (rays) in wireless locating finding systems. Overlapping fading multipath components for mobile-positioning are resolved by exploiting the fact that multipath components fade independently. Although fast channel fading is usually considered a challenge to the location finding process, it is used as an additional tool to detect and resolve overlapping multipath rays. A projection technique is also provided that exploits all possible a-priori channel information into a adaptive filtering algorithm, thus providing needed robustness to divergence of the adaptive algorithm that might result from possible severe data matrix ill-conditioning and high noise levels, which are common in wireless location applications.

The invention discloses a combination method based on evidence under a new evidence judgment standard. Through the method, the problems that in the prior art, evidence allocation is unreasonable, anda combination result goes against an actual result are solved. The method comprises the steps that 1, sensors are deployed, wherein sensor groups composed of multiple types of sensors are installed atcorners of a room, and the multiple types of sensors include temperature sensors, humidity sensors, smoke concentration sensors, carbon dioxide concentration sensors and infrared sensors; 2, a new evidence conflict judgment standard Pcor is confirmed, wherein evidence is determined according to data collected by the sensors, and the conflict degree among all the evidence is analyzed according tothe evidence; 3, weight coefficients of all the evidence are determined, wherein a support matrix is constructed according to conflict judgment standard Pcor values of all the evidence, trust degreesare confirmed according to the support matrix, and finally the trust degrees are used to confirm the weight coefficients of all the evidence; and 4, the evidence is combined, wherein basic probabilityassignment vectors are adjusted according to the weight coefficients, and combination is performed according to the DS evidence theory to obtain a more reasonable conclusion.

A method for the computer-based identification of mastitis of milk-producing animals of a herd, wherein individual animals are automatically identified by an animal identification means of a herd management system, and a milk sample assigned to the animal is taken and analyzed for the identification of mastitis. The method takes into account the fact that different animals in a herd have an individual tendency to contract mastitis, so that not all the animals in said herd have to be examined with the same frequency for a possible case of mastitis, and the method according to the invention correspondingly proposes that after the animal identification process, the herd management system decides whether a milk sample is taken and / or analyzed, based on animal-specific information that is stored in the herd management system.

We presented an approach for speeding-up image acquisition when tasked with localizing specific structures in FIB-SEM imagery. It exploits the fact that low-quality images can be acquired faster than higher-quality ones and yet be sufficient for inference purposes. We have demonstrated greater than five-fold speed-ups at very little loss in accuracy in the context of mitochondria and synapse detection. Furthermore, the algorithm we propose is generic and applicable to many imaging modalities that allow trading quality for speed.

An embodiment of this invention is about a method and a system for determining which line is causing noise based on peer feedback. Instead of relying on special monitors or algorithms to determine which line is causing the noise, the participants of the call are paired up and asked to identify themselves and to observe the other parties audio for line noise. One system capitalizes on the fact the ability of the caller to determine if the line is producing noise is in most cases far superior to that of the conference system. One method does not require evaluation to be done by the “controlling leg” (i.e. the moderator). That is, there is no need for a centralized authority which makes the decisions on line noise. Instead, each participant is asked to evaluate another participant.

The invention discloses a deep learning-based larceny auxiliary sentencing method, and belongs to the field of computers. The method mainly aims to solve the problem that the case semantic representation and the criminal period accurate prediction of a specific criminal name cannot be realized under the condition of less manual annotation, and comprises the following steps of defining the 11-dimensional features for formally describing a theft case from the perspectives of stolen article value, crime subject information, crime fact description and judgment result according to the criminal lawprovisions and sentences provisions about the theft and in combination with a disclosed theft first-review decision-making document; preprocessing the text of the judgment document; integrating into acorpus set, and training the word vectors; extracting the features except the values of the stolen articles and the criminal periods, and constructing a feature generator for each dimension of features by using a recurrent neural network so as to extract the feature values; using the linear regression and a multi-layer neural network model as the predictors, inputting the case feature vectors, and outputting the criminal period prediction results, so that the deep semantic understanding of the case can be realized and a clear criminal period prediction value can be given under the condition of less dependence on the manual annotation.

The present invention proposes a method for a dual-mode terminal to support an intelligent call, comprising the following steps: (a) the dual-mode terminal judges whether two networks have been opened, and if yes, executes the next step, otherwise, uses the opened network to call, and ends; ( b) The dual-mode terminal judges whether the called number is a mobile number, if yes, execute the next step, otherwise, press the default network or manually select the network call, and end; (c) determine the home network of the called number; (d) dual-mode The terminal judges whether the signal of the home network meets the requirement, and if so, calls through the home network. The present invention utilizes the fact that the mobile phone number segments belong to different network operators, intelligently judges the network ownership of the called number, and intelligently selects a network calling method that meets the user's requirements according to the signal conditions of different networks, thereby simplifying the The method that the user uses the dual-mode mobile phone to make a call makes the user not have to change the habit formed by using the single-mode mobile phone for a long time.

The computer implemented life form (CILF) is a belief program which excludes all three of the classical logic paradigms, it can then (at least for discussion purposes) be considered derived from a form of “non-classical” logic. Certainly, even the mere idea that reality itself could possibly be nothing more than a simulation, could easily be considered a new (neo-classical) and useful form of non-classical thought. The programming methods used by the CILF are non-computational, meaning they will not generate any independent fact or data. Instead, the CILF programming method will merely check data to form a state of belief or doubt upon the input data from which a new and improved data store can be more correctly and effectively established.

A production system of tense type is prepared for using databank to store data relating to problem to be solved, using rule set to present normal knowledge rule of problem field, using a reasoning machine to select and to use rules in said rule set as well as to operate databank according to reasoned result, containing tense attribute in both of fact and said rule.

The invention discloses a recurrent neural network sparse connection method based on block tensor decomposition. The recurrent neutral network sparse connection method is applied to the field of deeplearning neural network structure optimization and solves the problem that an existing method cannot perform analysis and optimization based on high-dimensional facts of input data and self redundancyof full connection and cannot quicken a training speed and maintaining or improve model accuracy at the same time. According to the method disclosed by the invention, sparsity of a full connection layer in a deep network is taken into consideration, an tensor decomposition idea is introduced, a sharing degree of network parameters is improved, a BPTT is utilized to perform model training, and themethod is adapted to most application scenes of an existing deep network; compared with an existing full-connection mode, a training speed and a convergence precision of the method disclosed by the invention are more greatly improved.

In this respect, the present arrangement provides a density current baffle that employs a baffle that is dimensioned to balance the fact that larger horizontal projections of density current baffles improve performance while simultaneously recognizing that an overly large baffle would simply not correctly function, be unnecessarily large and complex and may in fact have a negative impact on the otherwise ordinary operation of flows within the clarifier tank. Thus, the present arrangement provides a baffle that improves performance while satisfying the desire to remain conservative in defining an upper limit for the projection. The horizontal projection of the baffle is structured according to the following equation:HP=24″+0.4 inches / foot×(Diameter−30)

A knowledge automation engine to use in detecting product issues on products. A knowledge automation engine may evaluate a check against a fact to detect a product issue on a product and provide a user of the product remediation information. A check may contain a product issue description, a rule to evaluate against a fact in order to detect the product issue, and remediation information to help a user address the product issue if the product issue is detected on the product. Product issues may include product installation validation and known product bugs. Facts used by the knowledge automation engine may include product configuration facts.

A computer-based system and method is disclosed which constructs a personalized health risk assessment report by direct interactions between the individual and system that acquires pertinent and relevant personal information covering the blood type and Ayurvedic core constitution of a given individual. The system and method of the present invention enable an accurate personalized assessment to be generated for every individual interacting with the system. Once acquired, the facts of the individual's blood type and Ayurvedic core constitution are analyzed automatically by the database to identify a subset of potential health risks and weaknesses within the primary physical and emotional systems of the individual for each potential risk factor for disease that is uncovered in the database. Further, the automatically analyzed database of potential risk factors is used as the search tool for bringing to bear, on the discovery for each potential risk factor identified in each individual, a wealth of health and wellness knowledge that relates to and can be useful for correct and efficient recommendations for each of every individual's potential risks and weaknesses. This collection of information is analyzed to generate a final health risk assessment report of primary physical and emotional systems the individual may need to be aware of, potential risks and weaknesses within those identified systems, and dietary, nutritional and supplemental recommendations to bring into and maintain balance within those identified systems.

The invention relates to a design method of a high-fidelity and non-uniform shark-imitating groove microstructure of a large wind power generator blade, in particular to a method for performing differentiation design on the high-fidelity and non-uniform shark-imitating groove microstructures according to different radial region linear speeds of different fan blades so as to achieve efficient bionic resistance and noise reduction. The design method has the advantages that the fact that 'Reynolds number uses wing chord as characteristic length changes in the wind wheel radial direction' is fully considered when resistance and noise reduction structure design is performed on the large wind power generator blade, the method is scientific and reasonable compared with existing uniform design; the design method of the high-fidelity and non-uniform shark-imitating groove microstructure can ensure efficient actual using effect compared with existing abstract and simplified low-fidelity shark-imitating resistance and noise reduction structure design.

A method, computer program product, and computer system, for receiving a first set of ground truth instances from a first source. A second set of ground truth instances may be received from a second source. The first and second sets of ground truth instances may be weighed differently based on a level of trust associated with each of the first and second sources. The weighted first and second sets of ground truth instances may be applied in a machine learning task executed by a computer.