Abstract: Embodiments described herein provide for radiotherapy treatment plan analysis and guidance using voice commands and human language processing. A processor monitors a computer model generating a radiation therapy treatment plan comprising a projected dose-volume distribution for a structure of a patient or an attribute of the radiation therapy treatment plan. The processor receives a query indicating the structure of the patient. The processor generates a query term based on the received query. The processor retrieves data associated with the query term by querying data generated based on monitoring the computer model. The processor transmits the retrieved data to computer device.

Abstract: An electronic device and a method of controlling an electronic device are provided. The method includes obtaining an image comprising a text through a camera, identifying an input text, among texts included in the image, to be translated, obtaining a first vector corresponding to the input text by inputting the identified input text to an encoder of a translation model, identifying whether additional information is necessary to translate the input text by inputting the first vector to a first artificial intelligence model trained to translate the input text, based on identification that the additional information is necessary, identifying additional information among the at least one context information by inputting the first vector and at least one context information obtained from the image, and obtaining an output text corresponding to the input text by inputting the first vector and the identified additional information to a decoder of the translation model.

Abstract: Techniques for processing a read I/O operation that reads first content stored at a target logical address can include: determining, using the target logical address as a first key to index into a first cache, whether the first cache includes a first cache entry caching first metadata used to access a first physical storage location including the first content stored at the target logical address; responsive to determining the first cache includes the first cache entry, determining, using the first metadata as a second key to index into a second cache, whether the second cache includes a second cache entry caching the first content stored at the target logical address; and responsive to determining the second cache includes the second entry, returning the first content from the second entry of the second cache in response to the read I/O operation.

Abstract: Embodiments of the present disclosure provide a voice wake-up method, apparatus, electronic device, and readable storage medium. The method includes: acquiring a deep feature encoded by a voice activity detection (VAD) encoding layer through a VAD decoding layer, the deep feature being extracted from a digital audio signal used for VAD encoding and voice wake-up encoding; determining whether each audio segment in the digital audio signal is a voice according to a probability that each audio frame in the digital audio signal is a voice frame, each audio segment including a plurality of audio frames; and when determining that the audio segment is the voice, indicating, by the VAD decoding layer, to input a deep feature of the audio segment into a voice wake-up decoding layer to perform voice wake-up. The embodiments of the present disclosure improve the accuracy of voice wake-up.

Abstract: A computer-implemented method for managing data quality is provided. The method comprising determining, by a rule engine, a number of critical data points in a number of different software modules. A classifier is identified based on a data type of the critical data points, and the classifier is bound to the critical data points. The classifier scans the critical data points for anomality to verify an ability to correctly process the critical data points. A data quality report is generated based on the scan and displayed to an end user in a graphical user interface.

Abstract: An electronic device obtains a to-be-translated sentence. The electronic device divides the to-be-translated sentence into a preset quantity of clauses. The electronic device separately translates each of the clauses to obtain a respective translation result corresponding to each of the clauses. The electronic device combines the respective translation results corresponding to each of the clauses according to semantics to obtain a target translation sentence corresponding to the to-be-translated sentence.

Abstract: A hierarchical dictionary having methods of storing words based on frequency thereof in one or more documents which includes the steps of identifying a hash value corresponding to an inputted word; storing the word in a first hash map and in a second hash map having a substantially larger word storage capacity than the first hash map based on the identified hash value; clearing the first hash map at every predetermined period or triggering event; determining whether a frequency of the word as stored in the second hash map exceeds a predetermined value; and if so, promoting the word from the second hash map to a third hash map having a substantially larger word storage capacity than the second hash map for long-term storage and later retrieval.

Abstract: Embodiments described herein provide a pipelined natural language question answering system that improves a BERT-based system. Specifically, the natural language question answering system uses a pipeline of neural networks each trained to perform a particular task. The context selection network identifies premium context from context for the question. The question type network identifies the natural language question as a yes, no, or span question and a yes or no answer to the natural language question when the question is a yes or no question. The span extraction model determines an answer span to the natural language question when the question is a span question.

Abstract: A data processing method is described. The method includes acquiring a to-be-filtered dataset, the to-be-filtered dataset including a plurality of pieces of to-be-filtered source language data; filtering all source language data in the to-be-filtered dataset based on a target data filtering model to obtain target source language data remaining after the filtering, the target data filtering model being obtained through training performed by using a reinforcement learning algorithm; and acquiring markup language data corresponding to the obtained target source language data, and acquiring a machine translation model based on the target source language data and the acquired markup language data. In such a data processing process, a filtering rule in the target data filtering model is automatically learned by a machine in a reinforcement learning process. Apparatus and non-transitory computer-readable storage medium counterpart embodiments are also provided.

Abstract: Techniques for extracting and ranking multiple topic descriptions based on source contexts and subsequently selecting individual topic descriptions to surface based on recipient contexts. More specifically, a mining platform may extract, from a set of source documents making up a corpus, topic descriptions for various topics that are relevant to an enterprise. The mining platform may further rank the extracted topic descriptions based on a source context of those documents from which individual topic descriptions are extracted. Subsequently, when users access enterprise documents including term-usage instances of topics for which one or more topic descriptions have been extracted and ranked, a description serving module may select a topic description that is contextually appropriate for a recipient view the enterprise documents.

Abstract: An unsupervised text to speech system utilizing a lexicon to map input text to the phoneme sequence, which is expanded to the frame-level forced alignment with a speaker-dependent duration model. An alignment mapping module that converts the forced alignment to the unsupervised alignment (UA). Afterword, a Conditional Disentangled Sequential Variational Auto-encoder (C-DSVAE), serving as the self-supervised TTS AM, takes the predicted UA and a target speaker embedding to generate the mel spectrogram, which is ultimately converted to waveform with a neural vocoder.

Abstract: Techniques and solutions are described for facilitating the reuse of text translations. For a first instance of text in a human language, a first logical pointer value is assigned to the first instance of the text, pointing to a target having at least a first translation for the text. A second instance of the text is received. If only the first instance is present, the first logical pointer value is assigned to the second instance. If a third instance of the text is present, associated with a second translation and a second logical pointer value, user or process input is received determining whether the first logical pointer value or the second logical pointer value is assigned to the second instance. Context information can be provided to a user to help determine whether the second instance should be associated with the first translation or the second translation.

Abstract: Query execution on compressed in-memory data includes receiving, at a processor of an instance of a distributed in-memory database, a query for data from a table stored in the distributed in-memory database as compressed table data, obtaining results data responsive to the query from the table, and outputting the results data for presentation to a user. Obtaining results data includes allocating memory to identify allocated memory for decompressing the compressed table data, obtaining uncompressed table data by decompressing the compressed table data into the allocated memory, and obtaining the results data from the uncompressed table data. The allocated memory is deallocated in response to obtaining the results data. Compressing a table to form compressed table data is also described.

Abstract: A system includes a memory and a processor configured to detect that a primary mainframe job has failed, and extracts information related to the failed primary mainframe job, wherein the extracted information includes an error description of an error that caused the primary mainframe job to fail. The processor searches the memory based on the error description for a set of corrective actions corresponding to the error. In response to determining the set of corrective actions based on the search, the processor obtains a criticality index relating to the primary mainframe job. When the criticality index is lower than a criticality index threshold, the processor applies the set of corrective actions in relation to the failed primary mainframe job and processes the failed primary mainframe job after applying the set of corrective actions.

Abstract: Systems, methods, and computer readable media related to information retrieval. Some implementations are related to training and/or using a relevance model for information retrieval. The relevance model includes an input neural network model and a subsequent content neural network model. The input neural network model and the subsequent content neural network model can be separate, but trained and/or used cooperatively. The input neural network model and the subsequent content neural network model can be “separate” in that separate inputs are applied to the neural network models, and each of the neural network models is used to generate its own feature vector based on its applied input. A comparison of the feature vectors generated based on the separate network models can then be performed, where the comparison indicates relevance of the input applied to the input neural network model to the separate input applied to the subsequent content neural network model.

Abstract: A translating apparatus includes: a translator configured to translate an input sentence of a first language into a second language to generate a plurality of output sentences; a first dialogue manager configured to generate a meaning representation for the input sentence of the first language; a second dialogue manager configured to generate a meaning representation for each of the plurality of output sentences of the second language; and a determiner configured to determine a final output sentence among the plurality of output sentences of the second language based on the meaning representation for the input sentence of the first language and the meaning representation for each of the plurality of output sentences of the second language.

Abstract: Provided are computer implemented systems and methods for providing improved language translation, including translation of input text to an output translated text Selement. This may include providing for rule-based language translation. This may further include a plurality of machine translation models adapted to different translation domains. A machine translation selector is described that selects an in-domain machine translator from the plurality of machine translation models based on an input text element. The selected in-domain machine translator may generate a first translated text element from the input text element. A post-editor module may generate a second translated text element by predicting a post-edit to the first translated text element. The first translated text element and the second translated text element are evaluated by a quality evaluation model that determines a first metric associated with the first translated text element and a second metric associated with the second text element.

Abstract: The present disclosure provides a method and an apparatus of verifying information based on a voice interaction, a device, and a computer storage medium, and relates to a field of artificial intelligence technology. The present disclosure is implemented to include: acquiring a text of a voice response of a user to a voice inquiry, wherein the voice inquiry is provided for verifying information with the user; and inputting each character of the text of the voice response and a phonetic information associated with the each character to a pre-trained semantic analysis model so as to obtain an user intention information and/or an information of an object to be verified output by the pre-trained semantic analysis model, wherein the user intention information includes a confirmation, a denial, an answer, or a question.

Abstract: A printer-verifier is provided for printing and verifying a printed indicium. Printer-verifier includes imaging module for capturing image of printed indicium on print media, memory communicatively coupled to imaging module and configured to store print quality verification program, and processor communicatively coupled to memory. Processor is configured by print quality verification program to evaluate print quality of printed indicium, determine if print quality meets print quality standard, reject printed indicium on print media as rejected printed media when print quality does not meet print quality standard, and one of output location-specific feedback about rejected printed media or send a command signal to cutting device to destroy rejected printed media. Location-specific feedback is in same language as language associated with printer-verifier. A printer and methods for verifying the printed indicium are also provided.

Abstract: An embodiment of the present invention, a computer receives, on a first computer, a request to locate a topic. The computer determines a second computer to query for the topic. The computer queries the second computer for the topic. The computer receives identification information of participants of an instant messaging conversation corresponding to the topic. The computer stores, on the first computer, the identification information of the participants and indexing the stored identification information by the topic.

Abstract: A system for processing and presenting a conversation includes a sensor, a processor, and a presenter. The sensor is configured to capture an audio-form conversation. The processor is configured to automatically transform the audio-form conversation into a transformed conversation. The transformed conversation includes a synchronized text, wherein the synchronized text is synchronized with the audio-form conversation. The presenter is configured to present the transformed conversation including the synchronized text and the audio-form conversation. The presenter is further configured to present the transformed conversation to be navigable, searchable, assignable, editable, and shareable.

Abstract: A computer-implemented method for providing a voice interface system comprising: monitoring a first voice input during a training phase; monitoring manual inputs to a system during the monitoring the first voice input; storing training information associating the first voice input with the manual inputs; receiving a second voice input during an operational or runtime phase after the associating; identifying an instruction corresponding to the second voice input based on the training information; and executing the identified instruction.

Abstract: Interactive assistances for executing natural language queries to data sets may be performed. A natural language query may be received. Candidate entity linkages may be determined between an entity recognized in the natural language query and columns in data sets. The candidate linkages may be ranked according to confidence scores which may be evaluated to detect ambiguity for an entity linkage. Candidate entity linkages may be provided to a user via an interface to select an entity linkage to use as part of completing the natural language query.

Abstract: An object is to provide an interaction server in which inputs of an input keyword are reduced and the number of exchanges in interaction is reduced as a result. An interaction server 100 includes a truth table 104 configured to have question sentences including a sentence to be provided to a user and a plurality of searching keywords associated with, and a relational table 105 configured to have one main keyword and a plurality of related keywords related to the main keyword associated with each other. A search unit 102 acquires at least one of the one main keyword and any of the plurality of related keywords listed in the relational table 105 on the basis of an input keyword, further performs searching in the truth table 104 using the main keyword and/or the related keyword which has been acquired, and acquires a question sentence candidate which is response content corresponding to the main keyword or the related keyword.

Abstract: Machine learning translation models may be selectively tuned to provide custom machine translations. A request to translate input text from an input language to a target language may be received. A tuning data set for translating the input text to the target language may be identified and searched to select pairs of texts in the tuning data according to comparisons with the input text. A machine learning model used to translate into the target language may be tuned using only second texts in the target language in the selected pairs of texts. The tuned machine learning model may then be used to translate the input text into the target language.

Abstract: Automatic post-editing of machine translated content is disclosed herein. An example method includes presenting a user with machine translated segments of a document and their associated quality estimation scores, allowing a user to invoke an automated post-editing system for segments with unsatisfactory translation quality, inputting the segments into a generative AI model alongside contextual information, producing a revised translation of the segment using the generative AI model and iterating the generative AI process with varying prompts until a satisfactory translation is achieved or a predetermined number of attempts are reached.

Abstract: A method for translating a source text of a first language to a second language. The method includes receiving a translation request including the source text in the first language; selecting, from the source text, at least a first segment, associating at least one first metadata parameter with the first segment; providing the first segment to a first translation memory for determining a first set of translation proposals; determining a first quality score for each translation proposal; and comparing the first quality score of each translation proposal with a first predetermined acceptance threshold, and wherein based on the comparison, when a first quality score of at least one translation proposal is greater than the first predetermined acceptance threshold, the method comprises selecting a given translation proposal and providing the given translation proposal as an accepted translation of the first segment and as at least a part of an accepted translation of the source text.

Abstract: A system for processing and presenting a conversation includes a sensor, a processor, and a presenter. The sensor is configured to capture an audio-form conversation. The processor is configured to automatically transform the audio-form conversation into a transformed conversation. The transformed conversation includes a synchronized text, wherein the synchronized text is synchronized with the audio-form conversation. The presenter is configured to present the transformed conversation including the synchronized text and the audio-form conversation. The presenter is further configured to present the transformed conversation to be navigable, searchable, assignable, editable, and shareable.

Abstract: Apparatus and methods to decipher and analyze hidden/figurative meanings in communications using contextual analysis are provided. The apparatus and methods may include generating an initial list of slang words and phrases and their corresponding hidden/figurative meanings, searching the Internet for new slang words and phrases, determining the corresponding hidden/figurative meanings of the new slang words and phrases, and contextually analyzing a set of communications for hidden/figurative meaning. The apparatus and methods may include generating an alert if a relevant hidden/figurative meaning is found in the set of communications.

Abstract: Techniques described herein relate to facilitating end-to-end multilingual communications with automated assistants. In various implementations, speech recognition output may be generated based on voice input in a first language. A first language intent may be identified based on the speech recognition output and fulfilled in order to generate a first natural language output candidate in the first language. At least part of the speech recognition output may be translated to a second language to generate an at least partial translation, which may then be used to identify a second language intent that is fulfilled to generate a second natural language output candidate in the second language. Scores may be determined for the first and second natural language output candidates, and based on the scores, a natural language output may be selected for presentation.

Abstract: Systems and methods for generating phonetic spelling variations of a given word based on locale-specific pronunciations. A phoneme-letter density model may be configured to identify a phoneme sequence corresponding to an input word, and to identify all character sequences that may correspond to an input phoneme sequence and their respective probabilities. The phoneme-phoneme error model may be configured to identify locale-specific alternative phoneme sequences that may correspond to a given phoneme sequence, and their respective probabilities. Using these two models, a processing system may be configured to generate, for a given input word, a list of alternative character sequences that may correspond to the input word based on locale-specific pronunciations, and/or a probability distribution representing how likely each alternative character sequence is to correspond to the input word.

Abstract: [Problem] To provide a system capable of correcting an error in speech recognition by easier work than in the prior art. [Solution] A conversation analysis system comprising: a speech analysis unit 3 for analyzing contents contained in a conversation; a topics grasping unit 5 for grasping topics of the conversation; a topics-related correction terms storage unit 7 for, with respect to the respective topics, storing pronunciations to be speech-recognized or speech-recognized terms, and candidates for correction terms related to the pronunciations to be speech-recognized or the speech-recognized terms; and a conversation correction unit 9 for correcting the conversation analyzed by the speech analysis unit 3.

Abstract: A method, system, and computer program product for testing a container orchestration system are disclosed. The method includes replicating objects of a production cluster by extracting an object definition from an object and transforming the object definition to create a replicated object definition with an equivalent syntactic form. The replicated object definition requires fewer resources than the object definition. The method also includes applying the replicated objects of the production cluster to a simplified test cluster that replicates a configuration of the production cluster in a scaled down form. Additionally, the method includes testing, with the simplified test cluster, an upgraded version of the container orchestration system.

Abstract: Generate, for each of the words of a common vocabulary of first and second text corpora, a first word embedding vector in the first text corpus and a second word embedding vector in the second text corpus. Generate, for each word in a random sample of non-landmark words, an artificially shifted word embedding vector by modifying the first word embedding vector for that word. Train a machine learning classifier to predict whether an artificial shift has been injected for a given word, based on the artificially shifted word embedding vector and the second word embedding vector for the given word. Predict semantic shifts for at least a plurality of the words of the common vocabulary by providing the first word embedding vectors and the second word embedding vectors for at least the plurality of the words of the common vocabulary as input to the trained machine learning classifier.

Abstract: Aspects of the present disclosure include a system comprising a machine-readable storage medium storing at least one program and computer-implemented methods for detecting a language of a text string. Consistent with some embodiments, the method may include applying multiple language identification models to a text string. Each language identification model provides a predicted language of the text string and a confidence score associated with the predicted language. The method may further include weighting each associated confidence score based on historical performance of the corresponding language identification model in predicting languages of other text strings. The method may further include selecting a predicted language of the text string from among the multiple predicted languages provided by the multiple language identification models based on a result of the weighting of the confidence score associated with the particular predicted language.

Abstract: A created sentence evaluating device 1 using a neural network unit 10 of an encoder/decoder model in which an encoder unit 100 inputs a sentence in a first language, and a decoder unit 101 sequentially outputs word candidates for a sentence in a second language corresponding to the sentence in the first language and likelihood of the word candidates includes: an encoder input unit 13 configured to input a created sentence created in the second language to the encoder unit 100 sequentially for each word; and an evaluation unit 17 configured to evaluate words of the created sentence on the basis of word candidates in the second language and the likelihood of the word candidates output by the decoder unit 101 on the basis of an input from the encoder input unit 13.

Abstract: A method for creating a template for content including receiving at least one audience parameter representing at least one preference of a target audience when receiving the content, receiving at least one context parameter representing a context of the content, determining a content storyline based on the at least one audience parameter and the at least one context parameter, and generating a content template based on the content storyline.

Abstract: Provided are computer implemented systems and methods for providing improved language translation, including translation of input text to an output translated text element. This may include providing for rule-based language translation. This may further include a plurality of machine translation models adapted to different translation domains. A machine translation selector is described that selects an in-domain machine translator from the plurality of machine translation models based on an input text element. The selected in-domain machine translator may generate a first translated text element from the input text element. A post-editor module may generate a second translated text element by predicting a post-edit to the first translated text element. The first translated text element and the second translated text element are evaluated by a quality evaluation model that determines a first metric associated with the first translated text element and a second metric associated with the second text element.

Abstract: Various tools are disclosed for providing assistive or augmentative means to enhance the fluency and accuracy of persons having speech disabilities. These technologies may automatically ascertain and dynamically improve the accuracy with which automatic speech recognition (ASR) systems recognize utterances of persons having impaired speech conditions. In an embodiment, digitized audio information about a speaker’s utterance is processed to determine a set of candidate words matching the utterance. From these candidate words, a set of concepts is determined using a finite state machine model. A pictogram representing each concept is identified and presented to the speaker so that the speaker may select the pictogram corresponding to the best match of his or her intended meaning associated with the utterance. An action corresponding to speaker’s selection then may be performed. For example, displaying or synthesizing speech from textual information describing the selected concept.

Abstract: A system for testing the technical applicability of intellectual property rights having a computational linguistics module, which is designed to receive chains of technical graphemes and/or digitized symbolic images of intellectual property right specifications as input information, and to mechanically transfer the input information into an information model, an ontology memory, which is designed to store taxonomies, which are linked via inference rules, of conceptual classified object types, a semiotics module, which is coupled to the computational linguistics module and the ontology memory and which is designed to instantiate semiotic signs included in the information model, which is generated by the computational linguistics module, using conceptual classified object types stored in the ontology memory, and to create a contextualized information model, a construction module, which is coupled to the semiotics module and which is designed to generate a CAD model from the contextualized information model crea

Abstract: Interleaved conversation concept flow enhancement can include detecting a topic of an on-going conversation over an electronic communication system. In response to detecting a topic newly introduced into the on-going conversation, a topic-divergence metric (TDM) with respect to the original topic and the newly introduced topic can be determined. A topic-divergence response action can be initiated in response to the newly introduced topic diverging from the original topic by more than a predetermined threshold based on the TDM.

Abstract: In response to a user selecting a key on a keyboard in a first manner, a first alphanumeric character is displayed on a display device. In response to the user selecting the key on the keyboard in a second manner, a virtual key of a diacritic is displayed on the display device. In response to the user selecting the virtual key of the diacritic on the display device, the diacritic is displayed at a location of a second alphanumeric character on the display device.

Abstract: Disclosed are a method and system for verifying a resulting behavior of a graph query language. A lexical parser and a preset lexical rule are added to a BDD test framework, to construct an improved BDD test framework; a first Gherkin text in a graph query language standard is migrated to be a second Gherkin text in a graph query language; a first parsing result of the second Gherkin text is acquired from the improved BDD test framework; a first return result from a graph database is acquired, a comparison result between the first parsing result and the first return result is acquired, wherein the graph query language meets the graph query language standard if the first parsing result and the first return result are the same. The present disclosure solves the problem that the BDD test framework cannot recognize complex Gherkin texts, thereby facilitating development of graph databases.

Abstract: System and methods are provided for performing table navigation operations on a display, wherein the table comprises rows and columns and at least one cell that contains more information than presented on the display. Disclosed embodiments include receiving a scroll signal for scrolling the table, wherein the scroll signal results from a motion of a user on the display; in response to a vertical component in the motion, causing the table to scroll vertically; in response to a horizontal component in the motion, causing the table to scroll horizontally; receiving an enhancing scroll signal resulting from an enhancing motion for revealing hidden information within the at least one cell on the table; and in response to the enhancing scroll signal, revealing at least a portion of the hidden information within the at least one cell.

Abstract: Internet of Things (IoT) system and method of interfacing arbitrary non-network connected devices to wireless computer networks. The invention provides a configurable wireless communications module, in either fixed or removable formats, with wireless (e.g. WiFi) network connectivity. The invention uses at least one internal processor, which is configured to operate as a sandbox or virtual machine manner to isolate the code used to operate the arbitrary non-network connected device from the code used to operate the communications module.

Abstract: Techniques are described herein for subquery removal given two set operation-based subqueries in a query, where one subquery contains the result of the other. The described optimization technique of subquery removal is enabled by join and set operation-based containment of the set operation-based subqueries where semantic equivalence can be established for a given pair of set operation-based subqueries when some table(s)—with associated join condition(s), correlation condition(s), and/or filter predicate(s)—in one subquery are not considered. Subquery removal reduces multiple access to the same table and multiple evaluations of the same join conditions required to evaluate the query. When a subquery is removed from a disjunction, this may lead to other optimizations such as subquery unnesting, e.g., when the original query configuration would not permit query unnesting and the rewritten query (with one or more removed subqueries) permits unnesting.

Abstract: Method, apparatus and computer program product for querying a group-based communication platform are described herein. For example, the apparatus comprises at least one processor and at least one non-transitory memory including program code. The at least one non-transitory memory and the program code are configured to, with the at least one processor, cause the apparatus to at least store a plurality of electronic messages to a message corpus; receive a search query associated with the message corpus, wherein the search query comprises reaction metadata criteria; and apply the search query to the message corpus to generate a reaction search results set.

Abstract: A method includes obtaining n pairs of translation sentences of a source language and a target language, where each of the n pairs of translation sentences includes a source language sentence and a target language sentence that are translations of each other, extracting a source language segment from each source language sentence in the n pairs of translation sentences using an extraction rule of the source language, extracting a target language segment from each target language sentence in the n pairs of translation sentences, and generating an extraction rule of the target language based on n target language segments extracted from n target language sentences.

Abstract: Systems and methods are described for training an artificial intelligence model to infer a log sourcetype of a log. For example, logs may have different log sourcetypes, and logs having the same log sourcetypes may have different messagetypes. The artificial intelligence model may be a machine learning model, and can be trained using training data that includes logs with known log sourcetypes. Each log can be tokenized, filtered, converted into a vector, and applied to a machine learning model as an input to perform the training. The machine learning model may output an inferred log sourcetype, which can be compared with the known log sourcetype to update model parameters to improve the machine learning model accuracy. The trained machine learning model may be trained to infer a log sourcetype of a log regardless of the messagetype of the log.

Abstract: Methods, systems and computer readable media are provided for automatically creating a semantic model of a relational database for processing natural language queries. A computing device automatically extracts relational database metadata. The computing device prompts a user to enter textual labels for columns of the extracted metadata. The computing device automatically generates a schema annotation file based upon the relational database metadata and the textual labels for the columns. A natural language query is processed for the relational database using the schema annotation file.