41 results about "Semantic integrity" patented technology

The invention discloses a translation text integrity evaluation method based on bilingual text structure information. The method includes the following steps of firstly, extracting training corpus used for establishing a target end text unit integrity model from a target end texture structure tree; secondly, establishing the target end text unit integrity model through the training corpus generated in the first step; thirdly, integrating the target end text unit integrity model in a logarithm linear translation model, and conducting translation generation through an adaptability decoding method. According to the method, with the help of the bilingual text structure information, a decoder can further improve the translation quality of the current statistical machine translation by fully and approximately utilizing the semantic integrity information based on bilingual text language knowledge.

The invention discloses a voice processing method and device and equipment. The method comprises the steps that voice data of a to-be-processed voice are received, and the to-be-processed voice is thevoice transmitted by a first object in a several-round voice interaction process of the first object and a second object; according to the voice data, a confidence coefficient that the to-be-processed voice is a complete voice is determined by utilizing a semantic integrity model; and according to the confidence coefficient, mute waiting time for responding of the second object to the to-be-processed voice is determined. According to the method, based on the received voice data of the to-be-processed voice, the confidence coefficient that the to-be-processed voice is the complete voice is determined by utilizing the semantic integrity model, semantic incomplete voice information is effectively recognized, so that the mute waiting time for responding of the second object to the to-be-processed voice is dynamically adjusted according to the confidence coefficient, the situation that a user is interrupted when not finishing voice expression is avoided, and under the situation that semanteme of the voice is determined to be complete, mute duration is shortened, the interaction efficiency is improved, and the user experience is greatly improved.

The invention provides a speech VAD tail point determination method. The method comprises the following steps: receiving speech information of a user, dividing the speech information into data packets, and uploading the data packets to a server according to a time sequence; when it is judged that the current data packet is a mute packet, calculating the current mute duration t; and triggering theserver to detect the semantic integrity of the speech information according to the current mute duration t and a preset first threshold T1 to make the server determine a tail point of the speech information according to a semantic integrity detection result. The tail point of the speech VAD is not cut off by the intelligent equipment any more, the intelligent equipment uploads the data packet divided by the speech information to the server according to the time sequence, and the server is triggered to detect the semantic integrity of the speech information, so that the server determines the tail point of the speech information according to the semantic integrity detection result. The mute duration of the tail point of the speech information is changed into a dynamically adjustable value from the original fixed duration.

The invention discloses a Chinese word segmentation method and system based on word vector representation learning. The method comprises the steps of performing preliminary word segmentation processing on a text to be subjected to word segmentation to obtain a preliminary word segmentation text; inputting the preliminary word segmentation text into a BERT model for training to obtain corpus word vectors; inputting the corpus word vectors into a Bi-GRU model for training to obtain a plurality of feature word vectors; calculating cosine similarity of two adjacent feature word vectors to obtain acosine similarity value; judging whether the cosine similarity value is greater than or equal to a preset threshold value or not; and if the cosine similarity value is greater than or equal to the preset threshold, combining the preliminary word segmentation results of the words corresponding to the two adjacent feature word vectors. By adopting the method and the system provided by the invention, the problem of word segmentation of the unregistered words in the specific field is solved, the word segmentation performance is improved, the semantic integrity of the words is not damaged by the word segmentation result, and a large amount of manpower tagging corpora can be avoided.

The invention is applicable to the field of computer input methods and provides a method for constructing an input method word bank according to a Chinese language model. The method is composed of a Chinese language model module and a word creation module, wherein the Chinese language model module provides word formation information and word bank management information for the word creation module, the word creation module generates input method terms in batch by aid of database management software according to the word formation information provided by the Chinese language model, and addition / deletion, retrieval, ordering and other operation can be conveniently performed on the constructed word bank by use of the word bank management information provided by the Chinese language model module. The Chinese language model is customized according to the requirements of phonetic pause and semantic integrity in language communication and fully reflects characteristics of the Chinese language. Through the method, the key problems that traditional input method word banks are short of scientific word collection standards and low in word collection efficiency, word bank content is not comprehensive and not systematic and cannot be effectively managed, a user has no way of grasping the word bank content in an input process, the input speed is low, and language communication experience is lacked are effectively solved.

The invention discloses an image restoration method and system based on a progressive learning strategy, a storage medium and terminal device, and the method comprises the steps: inputting a to-be-restored image into a rough generation model, so as to obtain a rough generation result; extracting a feature map of the rough generation result to obtain a foreground feature map; and matching the foreground feature map with a background feature map, and reconstructing the foreground feature map by using the obtained background sample block with the highest matching degree with the foreground feature map. According to the method, when the large-area damaged area in the image is generated, the semantic integrity of the image is kept, clear details are generated, a good restoration effect is achieved on the image with the high structure, and therefore the image restoration experience of a user is improved.

The invention discloses a document theme partitioning method based on a domain knowledge map community structure, and the partitioning problem of document resources related to subject knowledge or document knowledge is mainly solved, so that documents related to a theme can be stored in a close logical place, and learning efficiency is improved. The document theme partitioning method is characterized in that a level community discovery algorithm based on the Fast Geedy algorithm and the GN algorithm is proposed, and a theme structure tree is built; in the process of feature extraction, knowledge units directly serve as feature vectors, and due to the fact that the knowledge units have semantic integrality, compared with a traditional method based on participles, the document theme partitioning method can reflect theme characteristics of the feature vectors better; in the process of calculating feature vector values, the method of combination of degree centrality and knowledge unit file frequency is proposed, wherein the concept of the degree centrality reflects the status of the knowledge units in a knowledge map whole situation. Through the method, accuracy of document theme partitioning is effectively improved, and the method is suitable for the document theme partitioning based on the domain knowledge map community structure in general scenes.

The invention relates to a method for ensuring semantic integrity for the isolation and control of TCP connection and a system thereof, which pertains to the field of information safety. After TCP processing step after the modification of the method of the invention ensures that an SYN message is received, the information of the characteristics of the message is notified to another network processing unit; the SYN message is regenerated and sent to a network by another network processing unit; after another network processing unit receives an ACK, a protocol stack generates a corresponding message. The system of the invention comprises a TCP handshake processing module, a TCP connecting and checking module, an application-layer data checking module, an application-layer protocol processing module and a safe data switching channel module. The method and system of the invention can not only isolate the TCP connection between an internal network and an external network and carry out application-layer data switching, but also can ensure the semantic integrity of the TCP and the situation that a server is closed, but a client still can establish the TCP connection cannot occur.