180 results about "Word generation" patented technology

The invention provides a method and system for automatically detecting semantic errors in a text. In order to detect the semantic errors in the text, the method generates a first set of features and a second set of features corresponding to each word in the text. A first set of features corresponds to statistical features of a word and a second set of features for a word corresponds to the features generated for a plurality of words surrounding the word in the text using matrix factorization techniques. The method, then, combines a first set of features and a second set of features to generate a third set of features corresponding to each word in the text. Thereafter, the method utilizes an Artificial Neural Network (ANN) model to detect semantic validity of each word based on a corresponding third set of features.

A system facilitates a search by a user. The system detects selection of one or more words in a document currently accessed by the user, generates a search query using the selected word(s), and retrieves a document based on the search query. When the document includes one or more links corresponding to a linked document, the system analyzes each of the links, prefetches the linked documents corresponding to a number of the links, and presents the document to the user. The system receives selection of one of the links and retrieves the linked document corresponding to the selected link. The system identifies one or more pieces of information in the retrieved document, determines a link to a related document for each of the identified pieces of information, and provides the determined links with the related document to the user.

A speech recognition dictionary creation device (10) that efficiently creates a speech recognition dictionary that enables even an abbreviated paraphrase of a word to be recognized with high recognition rate, the device including: a word division unit (2) that divides a recognition object made up of one or more words into constituent words; a mora string obtainment unit (3) that generates mora strings of the respective constituent words based on the readings of the respective divided constituent words; an abbreviated word generation rule storage unit (6) that stores a generation rule for generating an abbreviated word using moras; an abbreiivaed word generation unit (7) that generates candidate abbreviated words, each made up of one or more moras, by extracting moras from the mora strings of the respective constituent words and concatenating the extracted moras, and that generates an abbreviated word by applying the abbreviated word generation rule to such candidates; and a vocabulary storage unit (8) that stores, as the speech recognition dictionary, the generated abbreviated word together with its recognition object.

This invention relates to one digital keyboard multi-function and multi-language input system and method, which adopts Unicode to make each language character computer inner codes and output layer to get integration to realize input, translation and computation functions. The invention system comprises input module, judgment module, conversion module, word automatic cut module, prediction module, association module, translation module, computation module, telephone number module and word generation module connected by standard application programs.

A method and system for computational interpretation of natural language, wherein in an input string is received from input means. The input string is first tokenizde for providing a list of words. Then the list of words is stemmed for providing the words in the root form. The stemmed list is then tagged for providing classification tags for each word, which allows generating the context sensitive information for each word. Lastly said tags are used for parsing the structural dependencies for each word.

In an index generating device, a similarity calculating unit calculates access similarities indicating similarities of access histories between documents, based on history information indicating the access histories to the respective documents by users, and a similar document specifying unit specifies a similar document similar to a given document as a retrieval target, based on the access similarities or the like. A retrieval index generating unit generates a retrieval index for the given document as the retrieval target from words appearing in a document set consisting of the similar document and the given document as the retrieval target.

According to an embodiment of the invention, a method and apparatus are described for producing an error word for a data stream. According to an embodiment of the invention, a method includes receiving a plurality of data elements of a data stream; and updating an error word for the data stream by processing the plurality of data elements in a single iteration.

This invention relates to a unitary hand held medical prescription transcriber and printing unit. More specifically, a small, portable electronic device is provided which can record words spoken by a physician and from those words generate a printed medical prescription that is delivered directly from the device itself. The unit digitizes words spoken by the user, processes the speech to recognize the words spoken by comparing the acoustic spectral characteristics of the spoken words to the characteristics of a database of known words, formats those words in the form of a prescription, and prints the prescription onto paper using a small integral printer contained in the unit. In a alternate embodiment, the unit may display the prescription on a liquid crystal display screen and the user may edit the prescription before printing a hard copy. The unit assists in accurate dispensing of medicines by providing a legible prescription printout, while at the same time being neither time consuming nor difficult to operate.

The invention discloses a neural network question generation method based on answers and answer position information, and a neural network model is from a sequence to a sequence and consists of an encoder and a decoder. Wherein the encoder model mainly learns characteristics in an original text, and the decoder is used for generating a question sentence corresponding to an answer by using the characteristics. The complete generation process of the question sentence is divided into three modes, namely a question word generation mode, a dictionary word generation mode and a copy mode. Wherein the question word generation mode uses an answer-based model, participates in model calculation through context vectors irrelevant to position information of answers in an original text, and generates question words corresponding to the answers from a limited question word dictionary; Wherein the question word generation mode adopts context vectors related to position information of answers in an original text for calculation; The probability distribution of the copy mode directly uses an attention distribution related to location information of the answer in the original text.

The invention relates to a new word mining method and device, a computer device and a storage medium. The method comprises the following steps: obtaining text information to be subjected to new word mining; filtering the text information according to a preset filtering method to generate a plurality of text statements; inputting the plurality of text statements into a preset Nago algorithm model to generate a plurality of candidate words; inputting each candidate word into a pre-trained classifier, and carrying out word classification discrimination; and selecting new words meeting requirements according to a word classification judgment result. According to the new word mining method, the candidate words are generated by adopting the Nago algorithm, and the candidate words are detected and judged by adopting the pre-trained classifier, so that the words with low accuracy can be removed, and the new word generation accuracy is improved.

The present invention provides a voice recognition apparatus and a voice recognition method. An extraction unit respectively extractes feature amounts from a plurality of voice data relative to a plurality of rejected words when the rejected words are applied to recognition words registered in the word dictionary. A collation unit calculates a degree of similarity between the voice signal and each of the recognition words and reject words stored in the word dictionary by using the feature amount extracted by the extraction unit. A result determination unit determines whether or not a word having a high calculated degree of similarity corresponds to a reject word, when the word is determined as the reject word. A reject word generation unit generates an acoustic model from a voice signal having a high degree of similarity to the recognition word to which the reject word is added, calculated by the collation unit, and stores the generated acoustic model in the word dictionary as a reject word relating to the recognition word.

The embodiment of the invention discloses a model training method, an information generation method and device, electronic equipment and a computer readable medium. The embodiment of the model training method comprises the following steps: extracting a sample text and a tag word sequence of the sample text from a sample set; Determining a first word vector sequence corresponding to the sample textand a second word vector sequence corresponding to the tag word sequence; Taking the first word vector sequence as the input of a pre-established first initial model; taking the output of the first initial model and the first word vector sequence as the input of a pre-established second initial model, taking the second word vector sequence as the output of the second initial model, and training the first initial model and the second initial model by using a machine learning method; And determining the trained first initial model as a semantic vector generation model, and determining the trained second initial model as a tag word generation model. According to the embodiment, the tag word generation accuracy is improved.

The invention relates to the technical field of test and measurement, especially to a method and device for changing the phase jump of PSK. The device comprises a serial-to-parallel converter, a starting unit, a PSK word generation unit and a signal generator, wherein the serial-to-parallel converter is used for modulating a to-be-modulated baseband code element in a binary sequence and generating a modulation code element; the starting unit emits a phase jump signal when the modulation code element changes; the PSK word generation unit is used for generating PSK words in a transitional area within a period tau via a continuous function according to the phase jump signal; and the signal generator is used for generating a carrier wave whose phase continuously changes according to the PSK words. According to the method and device, the required phase shift is modulated, the transitional area ensures that the phase shift is continuous and smooth, thus, phase jump in the prior art is changed, and harmonic-wave components generated by phase jump is greatly inhibited.

The invention discloses a frequency-locked loop comprising a digital loop filter, a data collection unit, a comparison unit, an aging correction word generation unit and a correcting unit. The data collection unit collects digital data output by the digital loop filter, carries out the weight processing on the digital data to obtain a weight value, and provides the weight value as a current weight value to the comparison unit; and the comparison unit compares the current weight value with a standard weight value and provides a comparison result to the aging correction word generation unit which regulates a current aging correction word according to the comparison result to generate the latest aging correction word. When an aging frequency drift occurs in a crystal oscillator, the influence to the output frequency can be eliminated or decreased through the drift compensation of an aging compensation feedback loop.

The invention discloses an Indonesian-Chinese cross-linguistic retrieval method and system capable of integrating an association pattern with user feedback. The method comprises the following steps: utilizing a machine translation module to translate an Indonesian user query into a Chinese query, and submitting the Chinese query to a search engine module to carry out retrieval to obtain an initial retrieval result document set; utilizing the relevant feedback information extraction module of a user clicking behavior to obtain a user feedback initial retrieval related document set; carrying out preprocessing through a document preprocessing module to obtain an initial retrieval related document database; calling an all-weighted association rule mining module to construct an all-weighted association rule base; utilizing a cross-linguistic query expansion word generation module to establish an expansion word bank; utilizing a cross-linguistic query expansion implementation module to submit a new combined query to the search engine module to obtain a final retrieval result Chinese document; and utilizing a final result display module to submit a final retrieval result to the machine translation module, translating the final retrieval result into an Indonesian document, and returning the Indonesian document to the user. By use of the method and the system, cross-linguistic retrieval performance can be effectively enhanced and improved, and the method and the system have a good practical application value and popularization prospect.