288 results about "Content determination" patented technology

Systems, methods, and computer-readable media provide a context-specific prompt to answer a user query. The systems, methods, and computer-readable media determine a context based on content of a natural language request and / or determine a role of a user who submitted the natural language request. Additionally or alternatively, templates or RAG sources that will be used for prompt generation may include financials domain-specific knowledge or other domain-specific knowledge or insights. Inclusion of this additional information in the prompt enhances the context to promote more accurate results from a large language model. In one embodiment, the prompt templates are created from various RAG sources, such as payables, general ledger, receivables, and asset management, containing structured data and information specific to the financial domain, enterprise, or other domain, which helps craft accurate prompts. A prompt is generated that identifies a subset of available fields and other selected information based on the role or other context. The prompt template may contain domain-specific knowledge uses a relevant domain or enterprise information to drive relevant results, and an executable query is generated by a large language model based on the prompt. The executable query causes data to be retrieved from a database to generate a result, and information is displayed based at least in part on the result.

A system, method, and program product are presented for causing a processor of a computing device to receive content, determine a context of the content, and when the determined context is relevant for obfuscation, scan the content to identify information that matches, within the context, one of a plurality of data patterns indicative of being sensitive information. The processor further executes to assign a confidence value to the identifying process. When the sensitive information is identified with the confidence value above a confidence threshold, the processor executes to generate an obfuscated version of the sensitive information within the content, replace the sensitive information with the obfuscated version to form updated content, and output the updated content. The context may include at least one of a feature or a subject, a manner of use, and / or a determination of a user accessing, transmitting, or receiving the received content.

The embodiment of the invention relates to a document translation processing method and device, equipment and a storage medium, and the method comprises the steps: obtaining a to-be-translated document, segmenting the to-be-translated document into a plurality of fragments, and obtaining the context content of a first fragment in the to-be-translated document, the context content comprises original text content and translated text content of a first preset number of text units adjacent to a first fragment in front of the to-be-translated document, inputting the context content and the original text content of the first fragment into a translation model, translating by the translation model, and outputting the translated text content of the first fragment, and determining a translation result document based on the translation content of the first fragment. The context content of the original text content and the translation content containing the first preset number of text units adjacent to the first fragment is supported to be obtained, the context content and the original text content of the first fragment are input into the translation model, and the first fragment is translated by combining the context content to obtain the translation content of the first fragment. And the translation consistency of translations is improved.

The invention provides a search engine advertisement display method and a related device, which are applied to terminal equipment or a server, and the method comprises the following steps: determining the correlation between search content input by a user and one or more groups of attribute information, the one or more groups of attribute information corresponding to a first advertisement, each group of attribute information comprises at least one of the following items: a field related to the first advertisement, a descriptive statement for the first advertisement or a keyword of the first advertisement, and the first advertisement is an advertisement to be displayed determined based on the search content; and determining a target text for display based on the correlation, wherein the target text is a language description of the first advertisement. In the embodiment of the invention, the target text based on the advertisement can enable the user to quickly know that the pushed advertisement is related to the search content input by the user, so that the user feels that the pushed advertisement is possibly helpful to the user, the user can be better attracted, and the click rate of the pushed advertisement can be improved.

The invention relates to a fingerprint establishment method, a detection method and a content determination method of a traditional Chinese medicine composition mild drink. The traditional Chinese medicine composition comprises the following raw materials: lily, pawpaw, Chinese yam, poria cocos, cinnamon, orange peel, rhizoma polygonati, radix polygonati officinalis and mulberry. The fingerprint establishment method comprises the following steps: preparing a reference solution and a test solution; detecting the reference substance solution and the test solution by high performance liquid chromatography; a mobile phase adopted by the high performance liquid chromatography is an acetonitrile and formic acid aqueous solution, and the elution mode is gradient elution. According to the fingerprint spectrum of the traditional Chinese medicine composition and the content determination method, the traditional Chinese medicine composition can be effectively represented, the content of various components can be simply, conveniently and rapidly determined, and richer chemical information and more comprehensive basis are provided for quality evaluation of the traditional Chinese medicine composition.

The embodiment of the invention provides a task processing method and device.The method comprises the steps that a target to-be-detected text of a target task is obtained, the target to-be-detected text comprises at least one set of to-be-detected content, and the to-be-detected content comprises original content, content identification of the original content and at least one kind of translation content; determining the language type of each translation content; generating a prompt instruction corresponding to the to-be-detected content according to the content identifier, the original content, the translated content and the language type; and calling the large language model, and obtaining a task processing result corresponding to the target task based on the prompt instruction. By obtaining the target to-be-detected text, the prompt instruction enabling the large language model to execute the detection task can be generated for the at least one group of to-be-detected contents, so that the large language model can perform quality detection on the translation contents of the original contents based on the original contents in the text, the detection efficiency of the translation quality can be improved, and the user experience is improved. Therefore, the accuracy of translated contents is improved.

The invention relates to a fingerprint establishment method, a detection method and a content determination method of a traditional Chinese medicine composition fingered citron beverage. The traditional Chinese medicine composition is prepared from the following medicinal raw materials: fingered citron, orange peel, malt, liquorice, Chinese date, purple perilla, citron, almond and donkey-hide gelatin. The fingerprint establishment method comprises the following steps: preparing a reference solution and a test solution; detecting the reference substance solution and the test solution by high performance liquid chromatography; a mobile phase adopted by the high performance liquid chromatography is an acetonitrile and formic acid aqueous solution, and the elution mode is gradient elution. According to the fingerprint spectrum of the traditional Chinese medicine composition and the content determination method, the traditional Chinese medicine composition can be effectively represented, the content of various components can be simply, conveniently and rapidly determined, and richer chemical information and more comprehensive basis are provided for quality evaluation of the traditional Chinese medicine composition.

The invention relates to a content retrieval method and device, computer equipment, a storage medium and a program product. The method comprises the steps of obtaining a keyword of a target problem; according to the keyword, performing retrieval processing in a pre-established knowledge base constructed by at least one knowledge data in a text, a table and a picture through an index nested structure to obtain target retrieval content; and determining target reply content according to the target retrieval content. By obtaining the keyword of the target question and performing retrieval in the knowledge base with the index nested structure, the target retrieval content highly related to the question can be quickly and accurately positioned. In this way, blind search in a large amount of unorganized data is avoided, and the efficiency and accuracy of question answering are greatly improved.

The invention belongs to the technical field of quality control of traditional Chinese medicines, and particularly relates to an establishment method of an HPLC (High Performance Liquid Chromatography) fingerprint spectrum of a kidney-tonifying and liver-nourishing mixture and a multi-index component content determination method. The establishment method of the fingerprint spectrum comprises the following steps: mixing the kidney-tonifying and liver-nourishing mixture and a methanol aqueous solution, carrying out ultrasonic treatment, carrying out HPLC detection on an obtained test solution and a mixed reference substance solution, and introducing an obtained chromatogram into fingerprint spectrum software for treatment, so as to obtain the HPLC fingerprint spectrum of the kidney-tonifying and liver-nourishing mixture. The pretreatment method disclosed by the invention is simple, the characteristic components are completely reserved, and the high performance liquid chromatography is adopted, so that the precision is high, the reproducibility is good, and the stability is good; according to the fingerprint spectrum obtained by the invention, 15 common peaks are determined, the quality of the kidney-tonifying and liver-nourishing mixture can be effectively monitored, and a guarantee is provided for comprehensive and effective control of the quality of the kidney-tonifying and liver-nourishing mixture.

The invention provides a recommendation method and device, equipment and a medium, and relates to the technical field of artificial intelligence, in particular to the technical field of intelligent recommendation and software application. According to the implementation scheme, based on first interaction behavior information of a target user for current display content, whether the current display content can meet the interest of the target user or not is determined, and the current display content is determined from a candidate content library comprising a plurality of candidate content categories; in response to determining that the current display content does not meet the interest of the target user, obtaining an interest content category of the target user, the interest content category being determined from a plurality of candidate content categories based on content category information of a plurality of historical browsing contents of the target user; determining recommended content based on a plurality of candidate contents corresponding to the interest content category; and displaying the recommended content to the target user.

The invention relates to the technical field of digital management, and particularly discloses a publishing house full-process digital management system based on big data analysis, which comprises a publishing house topic selection planning module, a topic selection content determination module, a publishing house order management module, a book inventory monitoring module, a book quality evaluation module and a book copyright management module. According to the method, a topic selection feasibility evaluation model is established through book multi-source data fusion, topic selection accuracy is improved, production and printing management is intelligently planned through a paper format optimization algorithm, a book inventory prediction model is established through fusion of sales and inventory data, inventory allocation management is realized, and the production and printing efficiency is improved. The printing quality problem is automatically identified through standardized deviation analysis, copyright key links are subjected to uplink evidence storage, a press is promoted to transform from experience driving to data driving through a big data analysis technology, and scientization of topic selection decision, refinement of production management, standardization of quality management and control and full chain of copyright protection are achieved.

The invention relates to the technical field of traditional Chinese medicine analysis, in particular to a method for determining the content of five flavonoid components in a hypericum perforatum medicinal material. The content determination method comprises the following steps: preparing rutin, hyperoside, isoquercitrin, quercitrin and quercetin into a mixed reference solution; extracting a hypericum perforatum medicinal material by using an extraction solvent to prepare a test solution; the mixed reference substance solutions and the test solution with different volumes are subjected to high performance liquid chromatography respectively, a mobile phase A is acetonitrile, a mobile phase B is a phosphoric acid aqueous solution with the volume concentration of 0.08-0.12%, and elution gradient is adopted. The content determination method has the advantages of high sensitivity, simplicity and convenience in operation, low cost and high detection speed.

The invention discloses a preparation process of a Qingjin phlegm-reducing decoction material standard, an HPLC content determination method and a fingerprint spectrum establishment method aiming at the material standard, under the content determination method, the peak shapes of chromatographic peaks of all components are symmetrical, the separation effect is remarkable, and excellent stability, specificity, linear response, repeatability, precision and accuracy are shown. According to the present invention, the fingerprint characteristics are accurately recorded, the reversed-phase high performance liquid chromatography technology is adopted to successfully obtain 13 common peaks, the baicalin is adopted as the reference peak, the relative retention time and the relative peak area of the common peaks are calculated, and the reference substance comparison is further performed to identify the 8 common peaks; the established fingerprint spectrum method also has the characteristics of good chromatographic peak shape, high separation degree, stable method, excellent repeatability, high precision and the like. In conclusion, the invention provides a set of comprehensive and reliable Qingjin phlegm-reducing decoction material standard quality control reference system, and a scientific basis is provided for the overall quality evaluation of the Qingjin phlegm-reducing decoction.

The embodiment of the invention provides an evaluation content determination method and device and an evaluation summary information generation method and device, and the evaluation content determination method can display the evaluation summary information generated based on the existing evaluation content on an evaluation page of a target commodity in order to improve the decision-making efficiency when a user purchases the commodity. The evaluation summary information comes from key evaluation points of other buyers for the commodity, and can help the user to quickly know key advantages and disadvantages of the commodity. The target keyword contained in the evaluation summary allows the user to automatically screen out related specific evaluation content after clicking, and key information can be obtained without reading all evaluations one by one. According to the mode, the shopping decision-making speed of the user is increased, and the user experience is improved by reducing decision-making fatigue caused by a large amount of evaluation. The overall shopping experience is optimized by displaying evaluation summary information and providing evaluation screening functions related to keywords.

A digital human generation method, a model training method and apparatus, a device, and a medium are provided. An implementation solution is: obtaining material content; determining a plurality of scenarios from the material content based on a pre-trained scenario division model, where each of the plurality of scenarios corresponds to a content fragment of the material content that has complete semantic information; and for each of the plurality of scenarios, determining, based on a corresponding content fragment, target content corresponding to the scenario; determining scenario label information of the scenario based on the corresponding target content; and configuring a digital human specific to the scenario based on the scenario label information.

The embodiment of the invention provides a content reasoning method and device, equipment and a medium. The method comprises the steps of obtaining a task description text associated with a reasoning task; based on the text key content in the task description text, determining a task cognition template adapted to the reasoning task, and based on multiple pieces of template chain information in the task cognition template, determining a reasoning path associated with the reasoning task; when the reasoning task is executed according to the reasoning path, mapping the text key content to analysis class chain information, performing reasoning analysis on the analysis class chain information mapped with the text key content to obtain a content reasoning result, and performing result verification on the content reasoning result through verification class chain information to obtain a verification result; and determining a task reasoning result for the reasoning task according to the content reasoning result and the verification result. By adopting the embodiment of the invention, the inference analysis capability can be improved from the source, and the result output accuracy can be improved through a result verification mode.

The invention discloses an analysis method for determining multiple components of white spirit by high performance liquid chromatography-tandem mass spectrometry, and relates to the technical field of white spirit analysis and detection, and the analysis method comprises the following steps: firstly, preparing an ethyl carbamate precursor substance standard stock solution and a mixed standard intermediate solution, and diluting the mixed standard intermediate solution step by step to obtain a mixed standard working solution; then determining chromatographic and mass spectrometry conditions, wherein BEH Ami chromatographic column is adopted as the chromatographic condition, and 0.1% formic acid water-acetonitrile is adopted as a mobile phase for gradient elution; a mass spectrum condition adopts a multi-reaction monitoring mode under an electrospray positive ion mode; drawing a quantitative calibration curve; finally, preparing a to-be-detected sample for content determination. According to the analysis method for determining the multiple components of the Baijiu through the high performance liquid chromatography-tandem mass spectrometry, through chromatographic condition optimization and mass spectrometry parameter innovation, the detection efficiency is improved, peak area data of urea, citrulline and arginine can be synchronously obtained through single sample injection, the analysis time is short, and quantification is accurate.

The invention belongs to the technical field of artificial intelligence, and relates to a content review method, which comprises the following steps: obtaining marketing content to be reviewed, and determining a data type of the marketing content; obtaining a target content classification model according to the data type; classifying the marketing content by adopting a target content classification model to obtain a marketing content type; obtaining a pre-constructed review knowledge base, and retrieving from the review knowledge base according to the marketing content type to obtain a target review basis; and inputting the marketing content and the target review basis into a content review model, and performing illegal content judgment on the marketing content according to the target review basis to obtain a content review result. The invention further provides a content review device, computer equipment and a storage medium. The method can be applied to a business management program system of financial insurance, and can reduce the manual review cost of marketing materials and shorten the review period.

The invention provides a fingerprint spectrum and content determination method of a calamus dizziness relieving preparation, and belongs to the technical field of traditional Chinese medicines. The construction method comprises the following steps: detecting a test solution containing the calamus wind-quenching preparation by utilizing high performance liquid chromatography, and generating a fingerprint spectrum according to a detection result; an adopted mobile phase A is acetonitrile, and a mobile phase B is a 0.1% phosphoric acid aqueous solution; and carrying out gradient elution by using the mobile phases A and B. The fingerprint spectrum can be used for simultaneously detecting various chemical components in the calamus parviflora preparation, is high in separation degree, and can systematically, stably and fully reflect the quality of a product, so that the quality of the calamus parviflora preparation can be effectively controlled. Meanwhile, the construction method is simple and efficient, the detection time can be shortened, the use of a solvent is reduced, the detection cost is further greatly reduced, and the method has high specificity and selectivity on analysis of unknown and known components and is more beneficial to actual production and application.

The invention discloses a video text recognition method based on semantic collaboration and representation mining, which comprises the following steps of: selecting video frames, and constructing a video text recognition data set; obtaining the position information and the detection score of the text through a text detection network; the recognition content and the semantic score of the text are obtained through a semantic collaborative classifier, and auxiliary supervision is carried out on the recognition content and the semantic score through a feature mining module; screening out a high-quality text instance by using a collaborative score obtained by weighted averaging of a detection score and an identification score, and aligning and fusing position information and semantic information of the high-quality text instance as tracking representation; and inputting the high-quality text region and the tracking representation into a text tracking module to obtain a tracking ID, and determining a text recognition result of the to-be-recognized video in combination with the detection position and the recognition content. According to the method, the small and weak text areas and the similar text areas in the video text can be accurately identified, and the robustness of the model to visual interference of the external environment is improved.

Implementations relate to generating, using a generative model (e.g., an LLM), generative model output that reflects a response that is responsive to content of user input and that is responsive to temporal characteristic(s) of providing the user input (e.g., typing speed(s)). Input event(s) that are performed by a user in providing the user input are determined, and temporal features associated with the user input are extracted from the determined input event(s). The generative model is trained to process a combined representation of a content embedding determined from content of the user input and a temporal encoding that encodes the temporal features associated with the user input, in order to generate the response. The generated response thus includes content that varies even when two queries having the same word content are received, if input events for the two queries indicate different user intents.

Method, systems and computer programs for content provision are provided. A requestor node generates an anonymized and compressed representation of the user profile using a user data embedding machine-learning model inputting user data of a user profile. The content retrieval platform receives the anonymized and compressed representation of the user profile and a content retrieval request and inputs the anonymized and compressed representation of the user profile and the content retrieval request to a content determination machine-learning model to determine content in response to the content retrieval request.

The invention relates to the technical field of data processing, and discloses a processing method for realizing multi-modal data auditing based on machine learning, which comprises the following steps: identifying the content of a to-be-audited sample, determining a sample format and a data processing mode, and extracting different format features in the to-be-audited sample to construct a feature data group; determining a risk coefficient of the to-be-audited sample according to the feature data set; obtaining the position of the audio or video in the to-be-audited sample, determining a matching range according to the position, obtaining the content matching degree of the text and the audio or video in the matching range, and judging whether to adjust the risk coefficient or not according to the content matching degree; and comparing the final risk coefficient with a coefficient threshold, and determining the processing level of the to-be-audited sample according to a comparison result. According to the method and the device, unified auditing processing of various formats of data such as texts, audios and videos is realized, and the accuracy of cross-modal violation recognition is improved.

The embodiment of the invention discloses a large-model-based government affair official document intelligent generation method and device and a medium, and relates to the technical field of large models, and the method comprises the following steps: receiving an official document generation request input by a user, carrying out semantic analysis on the official document generation request, complementing missing information by querying a knowledge base and historical data, constructing a context framework, and generating a government affair official document; the official document generation request comprises an official document type, key information and a document issuing department, and the context framework comprises an official document structure template and filled entity content; generating manuscript content based on a context framework through a large language model, determining a plurality of candidate manuscript versions, and performing multi-dimensional scoring on each candidate manuscript version to determine a target manuscript version; and retrieving an official document format template corresponding to the official document generation request from a preset template library to perform semantic alignment and intelligent auditing on the target manuscript version and the official document format template so as to determine a target official document.

A system for presenting content to a user of a processing device during a journey, the system comprising a journey identification unit operable to identify one or more characteristics of the journey, a content determination unit operable to select and / or modify content in dependence upon the identified characteristics, and a content presentation unit operable to present the selected and / or modified content to the user of the processing device during the journey.

Techniques for training a language model for code switching content are disclosed. Such techniques include, in some embodiments, generating a dataset, which includes identifying one or more portions within textual content in a first language, the identified one or more portions each including one or more of offensive content or non-offensive content; translating the identified one or more salient portions to a second language; and reintegrating the translated one or more portions into the textual content to generate code-switched textual content. In some cases, the textual content in the first language includes offensive content and non-offensive content, the identified one or more portions include the offensive content, and the translated one or more portions include a translated version of the offensive content. In some embodiments, the code-switched textual content is at least part of a synthetic dataset usable to train a language model, such as a multilingual classification model.

The invention discloses a data analysis method and device based on a large model, a medium, equipment and a product, and relates to the field of large models, intelligent agents and data analys.The data analysis method comprises the steps that an intelligent interaction page associated with the intelligent agents is displayed; in response to a content input operation of the user in a natural language form in the intelligent interaction page, based on target content corresponding to the content input operation, determining a target data set required for performing data analysis on the target content, and obtaining metadata of the target data set; and calling a preset data analysis tool for data analysis based on the target content and metadata of the target data set through a large model to obtain an initial data analysis result, and outputting a target data analysis result at least based on the initial data analysis result. The conditions of model illusion and uncontrollable analysis results can be reduced, the stability of the analysis process is ensured, and the reliability and consistency of the analysis results are further improved.

Embodiments of the present disclosure provide a method and apparatus for generating code content. The method comprises: acquiring first code content generated by a first model on the basis of a code processing request; determining a first error associated with the first code content; acquiring reference information associated with the first error; and providing the code processing request, the reference information, and the first code content to a second model to generate second code content for the code processing request. In this way, the embodiments of the present disclosure can achieve construction of the reference information associated with the first error, so that the second model can generate the second code content for a first request on the basis of the reference information, thereby effectively improving generation quality of code content.

The invention discloses a test case generation method and device, equipment, a storage medium and a product. Identifying each text content in the system architecture diagram; determining a text content set according to the text contents; determining a transaction name set according to the transaction name of each transaction record in the key transaction list; according to the text content set, the transaction name set and a to-be-filled test case table, a target test case table is determined, and one test case in the target test case table is recorded as one target test case. According to the technical scheme, all the text contents in the system architecture diagram can be automatically recognized and arranged into the text content set, then the transaction names of all the transaction records are automatically extracted from the key transaction list, the transaction name set is formed, the text content set, the transaction name set and the to-be-filled test case table are combined, and the test case table to be filled can be filled. According to the method, related information is automatically filled, the target test case table is generated, the whole process does not need human participation, automatic generation of the test case is achieved, the influence of subjective experience of testers is effectively avoided, and the stability of test case generation is guaranteed.