11 results about "Sentence length" patented technology

The invention relates to a retrieval enhancement method and system based on key sentence vectors and a question answering system. The method comprises the following steps: firstly, cutting a plurality of texts into text blocks and extracting key sentences; converting the text blocks and the key sentences into vectors; further establishing a database, storing text blocks, text block vectors and key sentence vectors, and establishing an index relationship; when further retrieval is carried out, a retrieval text is converted into a text vector, approximate nearest neighbor search is carried out on the text vector and a text block vector in a database, and similar text blocks are found out; further obtaining text block vectors and key sentence vectors corresponding to the text blocks, and calculating the similarity with the text vectors; and finally, according to the text block vector similarity and the key sentence vector similarity, determining a target text block of the retrieval text. Due to the fact that the key sentence length is closer to the retrieval text length, vector similarity calculation is more accurate, the text block and the key sentence vector similarity are synthesized, the target text block can be effectively screened out, and efficient and accurate text retrieval is achieved.

This invention discloses an edge AI dialogue interaction adaptation and inference optimization method for children's smart hardware, specifically involving the field of edge AI dialogue interaction adaptation and inference optimization technology. The method includes acquiring the current round of child speech, corresponding text, user age information, historical interaction sequences, remaining memory, processor usage, remaining battery power, and network connectivity status at the edge computing terminal; performing child semantic normalization on the current round of child speech and corresponding text to extract sentence length, word difficulty, intent category, emotion category, and context depth to form a joint state value; mapping the child's age adaptation results, semantic carrying requirements, and edge terminal resource status together into an age semantic resource state graph; and performing local inference and out-of-bounds backoff after filtering and perturbing the target inference trajectory in this state graph.

This application provides a subtitle line-level alignment method, electronic device, and storage medium, including comparing the number of sentences in two text segments; intelligently segmenting the target language text according to its linguistic habits; using a pre-trained large-scale bilingual semantic embedding model to convert each sentence of the source and target language texts into a high-dimensional semantic vector; calculating the semantic similarity matrix between all sentence pairs and finding an alignment path with optimal semantic similarity globally; introducing a length ratio penalty factor to balance semantic similarity and sentence length reasonableness; if blank lines exist, locating the text line that produces the blank line and its adjacent areas, segmenting the preceding sentence of the target language text according to common punctuation marks of that language to organize sentence fragments and eliminate blank lines. This method adopts a "coarse-to-fine, progressively layered" processing strategy, ultimately outputting an alignment result without blank lines and with reasonable semantic matching.

The invention discloses a financial insurance knowledge base construction method and device, equipment and a medium, and belongs to the field of financial insurance. Dividing the document based on the sentence semantic similarity and a division granularity threshold to obtain a first document fragment, wherein the division granularity threshold changes along with the sentence length and the topic drift rate; based on the first document fragment, constructing a knowledge graph of multiple layers of themes including laws and regulations, terms, questions and answers and instances to obtain a financial insurance document knowledge base, wherein a current feature vector library comprises feature vectors of the first document fragment according to retrieval. According to the method and the device, the division granularity threshold is adaptively adjusted along with the sentence length and the topic drift rate, so that the semantic boundary robustness, the multi-level grading of the automatically updated knowledge graph and the semantic integrity of the retrieved information are ensured, and the technical problem that the retrieval precision is not high when the retrieval is carried out based on the existing financial insurance knowledge base is solved.

A computer-implemented method includes obtaining, from text corpus including article-summary pairs in a plurality of languages, a plurality of article-summary pairs in a target language among the plurality of languages, to form an article-summary pairs dataset in which each article corresponds to a summary; inputting articles from the article-summary pairs to a machine learning model; generating, by the machine learning model, embeddings for sentences of the articles; extracting, by the machine learning model, keywords from the articles with a probability that varies based on lengths of the sentences, respectively; outputting, by the machine learning model, the keywords; applying a maximal marginal relevance algorithm to the extracted keywords, to select relevant keywords; and generating a keyword-text pairs dataset that includes the relevant keywords and text from the articles, the text corresponding to the relevant keywords in each of keyword-text pairs of the keyword-text pairs dataset.

The invention discloses a multi-field machine translation method based on a large language model, and relates to natural language processing. The method comprises the following steps: constructing a multi-field monolingual data set, and dynamically generating a field term dictionary in combination with an expert term dictionary and a large language model self-reflection mechanism; on the basis of sentence length, syntactic complexity of SpaCy calculation and domain vocabulary proportion, screening single statements with different complexities, and prompting and guiding model learning by taking a term dictionary as a feature; a mixed course learning strategy is adopted, the difficulty degree of a data set is divided, monolingual continuous writing pre-adaptation and bilingual alignment fine tuning are alternately executed, and the multi-field adaptation capability of the model is improved in combination with progressive, multi-field mixed and monolingual and bilingual mixed training. According to the method, the BLEU and COMET indexes are improved by 10%-18%, the training time is shortened by 6-12 h compared with sequential fine tuning, small parameter quantity models such as Llama2-7B and the like are supported, and the method is suitable for multi-field machine translation scenes.

The invention relates to the technical field of propaganda and education content matching, in particular to a propaganda and education content matching system for analyzing the oral demand of a patient, and the system comprises an oral demand analysis module which is used for separating noun phrases and verb phrases in a text through word segmentation and part-of-speech tagging operation according to the original oral demand text of the patient, and outputting the separated noun phrases and verb phrases; and matching a preset symptom dictionary from the noun phrases to obtain symptom terms, and calculating the average sentence length of the text and the number of clauses to obtain a syntactic structure complexity value. According to the method, word segmentation and part-of-speech tagging are carried out on the original oral demand text of the patient, and linguistic features of the average sentence length, the number of clauses and the professional vocabulary frequency of the text are combined, so that the multi-dimensional initial state vector capable of reflecting the knowledge level, the focus of attention and the receiving capability of the patient is constructed, which can exceed simple keyword matching; the cognitive baseline of the patient is evaluated from the linguistic level, and deep understanding of the user state is achieved.

The application belongs to the technical field of natural language processing, and discloses a test reading comprehension answer sentence summarization method fusing CPT and an optimization model. The application summarizes and generalizes background material answer sentences based on a CPT model, and optimizes the text output by the CPT model by comprehensively using various factors such as semantic authority, question matching degree, dependency syntax and sentence length, to generate an answer sentence which is accurate in information, high in question matching degree, complete in syntax structure and summarizes and generalizes answer points, thereby improving the summarization and generalization capability of a question and answer type reading comprehension system. The achievement of the application can be applied to an education test tutoring platform to provide an effective technical solution for science and education.

A method may include receiving a first sequence of text comprising at least part of a sentence. A method may include predicting an end-of-sentence (EOS) by applying the first sequence of text to a first artificial intelligence (AI) model. The method may include determining, based on the predicted EOS whether the first sequence of text comprises a complete sentence, in response to determining that the first sequence of text comprises the complete sentence, providing text within the first sequence of text comprising the complete sentence as an input to a second AI model and in response to determining that the first sequence of text does not comprise the complete sentence, receiving a second sequence of text. The first AI model can be optimized based on a sentence length-based loss function.

Embodiments of the invention provide an object evaluation method and a text classification model training method. The object evaluation method comprises the steps of obtaining a spoken language text of a target object; classification prompt information and the oral language text are input into the text classification model, classification results corresponding to the multiple classification dimensions are obtained, the classification prompt information is used for guiding the text classification model to classify the oral language text from the multiple classification dimensions, and classification results corresponding to the multiple classification dimensions are obtained; the plurality of classification dimensions comprise at least two of a sentence length classification dimension, a grammar classification dimension and a syntactic structure classification dimension; and based on the classification result, generating an object evaluation result of the target object. By combining the classification prompt information designed for multiple linguistic classification dimensions, the text classification model is guided to perform multi-angle and fine-grained analysis on the oral text, so that a multi-dimensional classification result can be obtained; and then, based on a multi-dimensional classification result, an object evaluation result with high interpretability, result consistency and pertinence can be efficiently generated.

A method includes receiving a first sequence of text comprising a part of a sentence, providing the first sequence of text to a first artificial intelligence (AI) model to obtain an output, and predicting an end-of-sentence (EOS) based on the output. The method includes determining, based on the predicted EOS if the first sequence of text comprises a complete sentence, responsive to determining that the first sequence of text comprises the complete sentence, providing text within the first sequence of text comprising the complete sentence as an input to a second AI model and responsive to determining that the first sequence of text does not comprise the complete sentence, receiving a second sequence of text. The first AI model is optimized based on a sentence length-based loss function.