267 results about "Word type" patented technology

The invention relates to an intelligent error correction system of key words in the process of searching networks and a method thereof. On an Internet platform, firstly, a related linguistic model and a corresponding dictionary as well as a data index database are established through the training of related data information; secondly, a text is inputted, a Pinyin error correction part calculates the mistakes of Pinyin and characters, the error correction of characters is calculated by a fuzzy match; finally, all results are filtered according to the degree of association, a plurality of results are sorted to get the proximal results. The polyphone mistakes and character types as well as word types mistakes inputted by a user are corrected by means of the sound-character conversion and fuzzy error correction technical methods to correct the character replace mistakes, the unwanted character or the leakage of character mistakes, the character position mistakes, etc. in the input process. Moreover, the basic functions are expanded on the basis such as the English-Chinese and punctuations mixing error correction, the fuzzy match technique, the related prompt technique and the enhanced intelligence error correction.

The invention discloses a method and system for news event extraction based on neural network. The method comprises the steps of: conducting data pre-processing on original text of training corpus; introducing an event sentence sequence represented by a word vector into a bidirectional long and short memory network, using the bidirectional long and short memory network to train and obtain semantic features of each candidate trigger word; introducing the event sentence sequence represented by a word vector into a convolutional neural network, using the convolutional neural network to train and obtain global features of the event sentences where the candidate trigger words are in; according to the semantic features of candidate trigger words and the global features of the event sentences where the candidate trigger words are in, using softmax as a classifier to classify each candidate trigger word, and therefore finding out the trigger words for news events, and according to the trigger word type, judging the type of the event. The method and system can quickly and accurately extract news events and deal with news events contained in non-standard statements, and has the advantages of high efficiency and universal applicability.

A programming editor and creator allows users with minimal knowledge of a computer language syntax to generate computer source code using a structured rather than freeform process. A predefined command sentence is provided representing at least a portion of a programming command, and includes words, which may be defined as constant or enterable words. Enterable words may be required or optional, and correspond to entry components. Each word may be defined according to word types, including list words, variable list words, variable words, mathematical words, condition words or Boolean words. Enterable words may include repeatable words which may be inserted multiple times within a command to create conditional statements, mathematical equations and string concatenations. Input values are provided for the entry components. The predefined command sentence with the input values is converted to a completed command sentence corresponding to the programming command, which is translated to a target programming language.

A method is disclosed for evaluating a chat message sent between users of an online environment. The method may include associating each word in the chat message with metadata. The metadata identifies a word type and usage for each word in the chat message. This method may also include identifying one or more safety rules associated with the metadata. Each safety rule identifies an ordered sequence of one or more sets of words. This method may also include applying the safety rule to the chat message to determine whether a sequence of words in the chat message present in the ordered sequence of sets of words. Upon determining a word, from each set of words in the ordered sequence of sets of words, matches a respective one of the words in the chat message, the chat message is blocked from being sent to a message recipient.

The invention discloses a realizing method of sharing fonts, which arranges a shared dynamic database for storing individual font types in terminal units (100), and comprises the following steps: the terminal units (100) acquire fonts corresponding to the characters in the shared dynamic database arranged locally according to font information and characters in a document to be displayed, if fonts corresponding to the characters are acquired, the document is displayed according to the acquired fonts. The invention also discloses a system to share fonts and terminal units. The proposal of the invention can realize the separation between fonts and documents and can realize the individual document displayed in font by smaller font document volume.

The application discloses a voice-controlled remote control device, including a voice collection module used for collecting voice data, a voice identification module used for identifying the voice data into a text, a word splitting module used for splitting the identified text into one or more words, a matching module used for searching remote key words matched with words after word splitting, adjusting query priority of the remote key word type according to a matching result and / or adjusting the sequence of the remote key words in the remote key word type to which the remote key words belong, an instruction formation module used for obtaining one or more remote key words according to matching to form one or more remote control instructions and / or function switching instructions, and an emission module used for emitting the remote control instructions to electronic equipment using infrared and / or radio frequency signals. The application can reduce the matching time, and can perform customized self learning for different users so as to better adapt to user's usage habits.

Systems and methods are disclosed for a server learning new words generated by user client devices in a crowdsourced manner while maintaining local differential privacy of client devices. A client device can determine that a word typed on the client device is a new word that is not contained in a dictionary or asset catalog on the client device. New words can be grouped in classifications such as entertainment, health, finance, etc. A differential privacy system on the client device can comprise a privacy budget for each classification of new words. If there is privacy budget available for the classification, then one or more new terms in a classification can be sent to new term learning server, and the privacy budget for the classification reduced. The privacy budget can be periodically replenished.

Methods and apparatus are provided for generating aircraft cabin control commands from verbal speech onboard an aircraft. An audio command supplied to an audio input device is processed. Each word of the processed audio command is compared to words stored in a vocabulary map to determine a word type of each word. Each determined word type is processed to determine if an intent of the audio command is discernable. If the intent is discernable, an aircraft cabin control command is generated based on the discerned intent. If a partial intent is discernable, feedback is generated.

The invention relates to a method for the text introduction of a mobile terminal, which comprises the following steps of: starting the function of the text introduction; searching a prestored user introduction list, wherein user information of one or more users and introduction speed and introduction front information which are associated with the user information are stored in the prestored user introduction list; calling user introduction speed and front information which are associated with specific users; according to the introduction front information, adjusting fronts of characters displayed on a current page sequentially through the introduction speed; judging whether the introduction of the current page is completed or not according to the reading speed and the total amount of texts displayed on the current page; and acquiring corresponding data information of the next page automatically from the memory of the mobile terminal, and displaying page contents on a window interface of the mobile terminal. The invention also provides a method for determining the user introduction speed in the text introduction of the mobile terminal and a mobile terminal device with the text introduction.

The invention relates to an input method with an association function and can be applied to an electronic information processing unit to provide a text input function; the method is suitable for handheld devices and comprises the following steps: to provide a database and store multi-stroke words; when the user provides an input word, to compare the input word with the multi-stroke words; if the input word is corresponding to one of the multi-stroke words, then to select an associational word so as to rapidly finish the desired single word or words; when the associational word is selected, to switch the word to select the font configuration of the associational word depending on demand. If the input word is not corresponding to any of the multi-stroke words, the input word is stored in the database, and the arrangement order of the words in the database is adjusted according to the using frequency of the input word or the associational word, so as to achieve the purpose of inputting the text rapidly and accurately.

Systems and methods are provided herein for using a trained knowledge graph to accurately determine whether two phrases are conversationally connected. To this end, three phrases are received. The phrases are translated to strings of word types by determining what type of word each word of each phrase represents, and replacing each word of the first phrase with its respective type. A knowledge graph is then accessed to determine strengths of association between a combinations of each string of word types and any conversational category of a plurality of conversational categories. If any of the strengths of association exceed a threshold, the corresponding strings may be determined to be part of a conversation.

The embodiment of the invention discloses an event generation method based on text information, and the method comprises the steps: obtaining a to-be-processed text information set, and enabling the to-be-processed text information set to comprise N pieces of to-be-processed text information; grouping each piece of to-be-processed text information in the to-be-processed text information set to obtain M text groups, the to-be-processed text information in each text group corresponding to the same word type; carrying out clustering processing on the target text group to obtain a text result cluster, the target text group being any one text group in the M text groups, and each text result cluster comprising at least one keyword; and generating a corresponding first event according to the textresult cluster. The embodiment of the invention further provides an event generation device. According to the embodiment of the invention, on one hand, processing resources and computing resources are saved, the event mining efficiency is improved, and on the other hand, the grouped to-be-processed text information is clustered, so that the event generation accuracy can be effectively enhanced.

The invention provides a wall surface integrated interactive electronic white board, and belongs to the technical field of electronic information. The white board consists of a writing operation screen 1 and a large screen display panel 2, wherein the writing operation screen is positioned on one side of the large screen display panel and consists of a touch screen, a controller, a display, a large display screen or a projector, a projection curtain, a computer and a screen divider. When writing or operating in a near distance, a user uses the higher-resolution writing operation screen relevantly smaller than or as large as the large display panel, and students can watch the large display screen or a projection in a far distance, so the problem of height restraint of the user when writing operation is performed on an overlarge screen and the problem that the user who writes on a large screen in the near distance feels dazzling light and stops the sight line of the students and projection light are solved. The electronic white board combined with software realizes writing on a computer screen, and computer operation and word type-in are performed without a key and a mouse through a pen, so the right key function of the mouse is enhanced, and the left key and the right key of the mouse can be switched easily and quickly.

The invention discloses a domain concept extraction method based on Deep Learning. The method includes extracting samples in a training corpus, adopting word frequency, document frequency, inverse document frequency, word length, word frequency variance and domain consensus as feature vectors, training and acquiring a deep network model, which is capable of representing the complex mapping correspondence between the word-type filed concept multi-dimensional feature vectors and class labels, on the basis of the Deep Learning technology, and finally comparing the deep network model established on the basis of the Deep Learning technology, an optimized BP neural network model and mainstream KNN and SVM models in the testing step. According to the tests, the optimal test effect is acquired through the deep network model established on the basis of the Deep Learning technology.

The invention relates to a method and device for asking a question in a dialogue system, electronic equipment and a computer readable medium, and belongs to the technical field of computers. The method for asking the question in the dialogue system comprises the following steps that: receiving input information; through the coder of a model, coding the input information into a hidden state sequence; and inputting the hidden state sequence into the decoder of the model, and generating corresponding output information, wherein each word for forming the output information independently belongs toone of a preset quantity of pre-defined word types, and the output information relates to the input information.

An information storage medium on which video data is recorded, comprising: a plurality of clips as a recording unit in which the video data is stored; and text data for subtitles, which is recorded separately from the plurality of clips and can be shared with the Images of data, including data for providing subtitles in at least one language, are layered and subsequently outputtable.

The invention discloses a Uigur display method and a relevant device, in order to fulfill Uigur display. The invention belongs to the technical field of word edition in an interview and edition system for ethnic presses. The method comprises: Determine a character under display to be a Uigur character, and insert the Uigur character into a buffer memory; perform shape selection and adjust character width according to the location of the Uigur character, and then, perform image display for the buffered Uigur character in a device coordinate system in the opposite direction against X-axis in a client coordinate system according to the selected font and adjusted character width. Besides, the invention also provides a Uigur display device according to the method.

The invention provides a vehicle color recognition method and device, and the method comprises the steps: obtaining a vehicle color recognition region comprising a vehicle image from a vehicle sample image; dividing the vehicle color recognition region into eight subregions in the shape of the Chinese character Mi, and extracting the histogram of each subregion; enabling the statistical information represented by the histograms as a color feature input classifier, so as to obtain a color region result of the color recognition region. Because the histograms at least comprises a {H, S} two-dimensional histogram and an Sqrt(S*V) histogram, the method is good in color distinguishing characteristics, wherein the {H, S} two-dimensional histogram gives consideration to hue and saturation, and the Sqrt(S*V) histogram gives consideration to the saturation and brightness. Meanwhile, the method also obtains a part of position information besides the color information of the histogram of all subregions obtained through the Mi-shaped dividing. Therefore, the method can reduce the impact on color recognition from illumination change and the interference caused by posture changes of a vehicle, vehicle windows and the ground, and improves the accuracy of vehicle color recognition.

A programming editor and creator allows users with minimal knowledge of a computer language syntax to generate computer source code using a structured rather than freeform process. A predefined command sentence is provided representing at least a portion of a programming command, and includes words, which may be defined as constant or enterable words. Enterable words may be required or optional, and correspond to entry components. Each word may be defined according to word types, including list words, variable list words, variable words, mathematical words, condition words or Boolean words. Enterable words may include repeatable words which may be inserted multiple times within a command to create conditional statements, mathematical equations and string concatenations. Input values are provided for the entry components. The predefined command sentence with the input values is converted to a completed command sentence corresponding to the programming command, which is translated to a target programming language.

The invention provides a full automatic lamination device for a Chinese character "Ri" shaped type or E1 type iron core of a transformer. The full automatic lamination device at least comprises a lamination frame, a guide rail device and a screw rack which are fixed above the lamination frame and parallel to each other, a three-column lamination, a yoke column lamination and an upper yoke column lamination which move along the guide rail device and the screw rack; the three-column lamination, the yoke column lamination and the upper yoke column lamination are located on the same straight line and respectively connected with the guide rail device; the three-column lamination, the yoke column lamination and the upper yoke column lamination are respectively driven by a group of servo systems and a helical gear; the three-column lamination is provided with a transmission used for adjusting the edge distance of the transformer iron core; the three-column lamination, the yoke column lamination and the upper yoke column lamination are linearly stacked by means of intermediate alternating; left and right laminated reclaiming is adopted, the structure is compact and the efficiency is high, the lamination technological requirement on the Chinese character "Ri" shaped type or E1 type iron core of different transformer manufactures can be satisfied, and manual tedious labor can be replaced, thereby saving time and labor, realizing full automatic lamination and having strong universality.

A computer-based offline handwritten Uyghur word recognition method based on word segmentation belongs to the word processing technique of word pattern recognition. The method includes the following steps that firstly, according to a Uyghur rule and a morphological structure, a Uygur word character library is built and includes three types of graphemes, namely main bodies, additions and dots; secondly, by segmenting a word image, three grapheme sequences are obtained, and according to the graphemes, different feature extraction and classification devices are designed; finally, through a hierarchical matching model of the graphemes, conjoined segments and words is built through the Bayesian network, reasoning calculation is conducted on grapheme features to obtain the recognition confidencecoefficients of word types, grapheme recognition information and word-formation prior information are mixed, and a word recognition result is obtained. By means of the offline handwritten Uyghur wordrecognition method, Uyghur words which are written unrestrainedly, naturally and smoothly can be robustly recognized, the training type required by the algorithm is fixed, and the expansibility of the algorithm is very high.

A character identification method being optimized by template includes using at least one training character in training character set to structure a font, placing said font in different cluster set to pick up public templates of said font then storing picked up templates in template storing module, storing internal code of each font to be corresponded to pointer of various font public template to form font index table by template output unit, carrying out level by level matching of character to be identified with various public templates pointed by a found font index table then recording matched result for obtaining candidate character.

The invention provides a method for extracting road traffic information from an Internet unstructured text. The method comprises the steps of defining a data structure of the road traffic information and a description feature word type of the road traffic information, expanding a few manually-established basic extraction modes to obtain an extraction mode bank, generating a feature word type sequence after the input Internet unstructured text is preprocessed, obtaining a matched extraction mode of the input text according to the similarity of the feature word type sequence, utilizing the matching extraction mode for extracting a positioning information element and a type information element of the road traffic information from the Internet unstructured text, utilizing a regular expression and a judgment rule for extracting a time information element from the input text, and obtaining the road traffic information through the combination of the positioning information element, the type information element and the time information element. By the means of the method for extracting the road traffic information from the Internet unstructured text, real-time processing can be carried out on the unstructured text collected from the Internet, the road traffic information can be extracted, and the traffic information collecting means are enriched.

The invented system suitable for inputting single Chinese character comprises first input unit, second unit, speech recognition unit, font recognition unit, font library unit, Chinese character features processing unit and display unit. The method includes following procedures. The phonetic feature of Chinese characters is input by first input unit. The second input unit inputs N font features ofsingle character. The speech recognition unit and font recognition unit recognize speech feature and font feature. The Chinese character features processing unit combining the speech feature and fontfeature selects out a set of candidate characters, which accord with speech feature as well as accord with N font features, from font the library then outputs and displays the set.

The invention discloses a Chinese character 'sound shape' input method of a computer keyboard and a numerical keyboard; the Chinese pinyin alphabets are respectively defined in a cell phone keyboard and 8 number keys from 2 to 9 at a computer keypad or character keys of the computer keyboard; 5 basic strokes of the Chinese character font 'horizontal stroke, vertical stroke, left-falling stroke, point, turning stroke' are respectively defined in the cell phone keyboard and 5 number keys '1,2,3,4,5' at the computer keypad, or 5 character keys 'G,H,F,J,D' at the computer keyboard. The single Chinese character takes the number keys or the character keys which correspond to the first letter of pinyin symbols and the first two strokes and the last stroke of the font; and the two-word phrase takes the number keys or the character keys which correspond to the first letter of pinyin symbols of each Chinese character and the first stroke of the first Chinese character, the first stroke and the last stroke of the second Chinese character. For the invention, the repeated codes of the single Chinese character and the phrases are less, the input process is simplified, and the invention is simple to learn, and the input speed is fast.

The invention provides a Chinese font subset-based webpage font display method and system. According to the method, a user webpage is browsed, a js plug-in preset in the webpage runs, when text contents of the currently browsed webpage do not start to be loaded, the js plug-in obtains a network address of the currently browsed webpage, and calculation is performed based on the network address to obtain a file name of @font-face corresponding to the currently browsed webpage; and according to the file name of the @font-face, a @font-face file corresponding to the currently browsed webpage is obtained from a file storage space, if the @font-face file is obtained successfully, the @font-face file is dynamically cited to the currently browsed webpage, and a font effect generated by a @font-face statement in the @font-face file is applied. Font subsets corresponding to the webpage are stored in the file storage space in advance, so that when browsing the webpage again, a user calls a font in the font subsets directly from the file storage space and loads the font to the webpage for display; and therefore, the calculation amount of a subset server is reduced and the user experience is improved.

The invention discloses a scenic spot abnormal event extraction method based on a composite neural network. The method comprises the following steps of carrying out data preprocessing on an obtained event original text corpus, converting an event sentence into a word vector, transmitting a sequence of the word vector to a bidirectional long-short time memory network, and training by utilizing thebidirectional long-short time memory network to obtain semantic characteristics of each candidate trigger word; transmitting the event sentence sequence represented by the word vector into a convolutional neural network, and performing training by using the convolutional neural network to obtain the global features of the event sentence where the candidate trigger word is located. According to themethod, the semantic features of candidate trigger words and the global features of sentences where the candidate trigger words are located are synthesized, the softmax is used as a classifier to classify each candidate trigger word, so that the trigger words of scenic spot abnormal events are found out, and the event types are classified according to the manually marked trigger word types. According to the method, the scenic spot abnormal events can be quickly and accurately extracted, the abnormal events in complex redundant texts can be processed, the efficiency is high, and the universality is good.

A radio frequency receiver for receiving supplementary data transmitted with a radio station signal. The supplementary data includes one or more information word fields containing supplementary data transmitted with the radio station signal. The receiver includes a host processor for processing the supplementary data; a supplementary data demodulator and decoder for demodulating and decoding a received radio signal to generate a stream of bits representing one or more information word fields; an information word type memory for storing one or more information word type fields; detection circuitry arranged to receive and compare the information word type field decoded by the supplementary data decoder with one or more information word type fields stored in the information word type memory, and to provide a matched condition signal in response to a match between one or more decoded information word type fields and one or more of the corresponding stored information word types fields. The receiver is arranged to, in response to the matched condition signal, cause the corresponding supplementary data to be processed by the host processor.

Systems, devices, and methods of the present invention relate to text classification. A text classification system accesses an utterance of text. The utterance includes at least one word. The text classification system generates a parse tree for the utterance. The parse tree includes at least one terminal node with a word type. The terminal node represents a word of the utterance. The text classification system applies one or more rules to the text. The text classification system then classifies the utterance as a question or a request for an autonomous agent to perform an action.