Abstract: An artificial intelligence (AI) method includes a processor receiving, from a human user, a query on an external database; performing transformations on the query to produce a refined query; generating an embedding of the refined query; and generating a system prompt and a user prompt based on the refined query. The processor then applies the embedding to a vector database by executing a similarity routine to identify discrete vectors in the vector database most similar to the embedding, and collects the most similar vectors for application to a large language model. Next, the processor applies the most similar vectors, the refined query, and the system and user prompts to a large language model to generate a comprehensive response to the query. The response includes a text document that is generated by execution of the large language model.

Abstract: The present disclosure provides computing systems and methods for cataloging, retrieving, and/or organizing user-generated content associated with objects. Aspects of the disclosure are directed to a systems and methods which utilize computers to enable users to interact with libraries of user-generated content associated with cataloged objects. For example, a user can capture one or more images of a real-world object, label or otherwise annotate the object with various types of user-generated content and organize the object and its associated content into one or more libraries. The user-generated content can then be provided to other users upon the receipt of images of the same object or an object displaying similar features.

Abstract: In accordance with an embodiment, described herein is a system and method for loading and transforming data to a cloud database, using a hybrid set-based extract, load, transform approach. During the loading and transforming of large amounts of data, from one or more data sources, a data synchronization application sends small batches of data, in parallel streams, to the database, which operates as a transformation engine to load the data. Each stream can be associated with a separate staging table in the database. For each staging table, the subset of data therein is transformed and loaded, before the next subset is processed. A transform and merge process operating at the database can then be used to transform and merge the data, from each of the staging tables, to the target table.

Abstract: A building management system includes one or more memory devices configured to store instructions thereon, that, when executed by one or more processors, cause the one or more processors to receive a publication by an agent on an agent communication channel, the publication comprising timeseries data, identify, based on the publication, an object entity of an entity database associated with the agent, wherein the entity database includes one or more object entities and relationships between the one or more object entities and one or more data entities, identify a data entity related to the object entity based on a relationship of the relationships relating the object entity and the data entity, and ingest the timeseries data into the data entity.

Abstract: System and method for creating clips of larger audio streams based on very accurate in point and out point and for searching for the clips. The system and method use byte signatures in MP3 or media files to accurately and efficiently find a specific cue point, such as for example, the beginning of a sentence in a podcast. The system and method are configured to search and locate a given clip even if that sequence moves inside the stream, such as for example, if the audio file adds additional content like advertisements or other audio sequence, the hash representation of the cue point can still be found with millisecond accuracy.

Abstract: There is provided with an image searching apparatus. A searching unit searches a plurality of search target images for an image relating to a query image based on a first similarity to the query image by a first evaluation scheme. A first obtaining unit obtains a feedback indicating correct or incorrect regarding the searched image. A second obtaining unit obtains an evaluation of an effect of changing a similarity evaluation scheme to a second evaluation scheme from the first evaluation scheme in accordance with the feedback. A control unit controls, in accordance with the evaluation of the effect, a search of the plurality of search target images for the image relating to the query image based on a second similarity to the query image by the second evaluation scheme.

Abstract: Exemplary embodiments relate to a method for selecting an image of interest to construct a retrieval database including receiving an image captured by an imaging device, detecting an object of interest in the received image, selecting an image of interest based on at least one of complexity of the image in which the object of interest is detected and image quality of the object of interest, and storing information related to the image of interest in the retrieval database, and an image control system performing the same.

Abstract: Embodiments for providing an intelligent media data service in a computing environment by a processor. One or more sections of media data are identified and annotated (e.g., tagged) for a user based on a degree of relevancy between a user profile and the media data, wherein the media data include media classification, topic detection, speaker detection and noise detection. The one or more sections of media data are selected for the user based on the tagging of the or more sections.

Abstract: A computer system is provided, including a processor coupled to a mass storage device that stores instructions, which, upon execution by the processor, cause the processor to store an original string formed of a plurality of characters, perform a wordbreak algorithm on the original string, and tokenize the original string to generate a processed string including a plurality of word tokens separated by spaces. The processor is further configured to generate an offset map between locations within the word tokens in the processed string and corresponding locations in the original string and classify a portion of the processed string as a target. The processor is further configured to identify target characters in the original string that correspond to the target using the offset map and perform a predetermined action on the target characters in the original string.

Abstract: The present disclosure relates to an object selection system that automatically detects and selects objects in a digital image based on natural language-based inputs. For instance, the object selection system can utilize natural language processing tools to detect objects and their corresponding relationships within natural language object selection queries. For example, the object selection system can determine alternative object terms for unrecognized objects in a natural language object selection query. As another example, the object selection system can determine multiple types of relationships between objects in a natural language object selection query and utilize different object relationship models to select the requested query object.

Abstract: An example method of performing searches by comparing regions includes identifying a search point contained in geo-located data. The search point is nearest in distance to a query location. A search region associated with the search point is determined. The search region includes neighbor points contained in the geo-located data that are nearest the search point. Centroids corresponding respectively to the plurality of neighbor points of the search region are determined. The centroids of the search region are compared to other centroids corresponding to other regions of the geo-located data. The number of other centroids that match with the plurality of centroids are counted, and the counts are placed in descending order, to list the other regions having the highest similarity first. An ordered list of the other regions is generated based on their respective similarities to the query region.

Abstract: A method for generating a composite visibility indicator of at least one entity includes selecting a first category of entities; selecting a first position; associating the first position with a first geographical zone, said first geographical zone defining a zone around the first position as a function of the selected category; generating at least one request to collect a plurality of indicators, and generating a composite visibility indicator for each entity of the first category selected within the first zone, the visibility indicator being a function of the first indicator and of the second indicator.

Abstract: The present disclosure relates to a media retrieval method and apparatus, wherein the method comprising: acquiring a media feature of a query media as a first media feature which contains a plurality of first media feature units; performing, according to each individual of the first media feature unit, a first ranking on a plurality of existing media, and taking, according to a result of the first ranking, first k existing media as a first candidate media set, where k is a positive integer; and performing, according to the plurality of the first media feature units arranged in sequence, a second ranking on the first candidate media set, and taking, according to a result of the second ranking, first n first candidate media as a retrieval result, where n is a positive integer.

Abstract: The present disclosure provides systems and methods that include or otherwise leverage use of a multiscale quantization model that is configured to provide a quantized dataset. In particular, the multiscale quantization model can receive and perform vector quantization of a first dataset. The multiscale quantization model can generate a residual dataset based at least in part on a result of the vector quantization. The multiscale quantization model can apply a rotation matrix to the residual dataset to generate a rotated residual dataset that includes a plurality of rotated residuals. The multiscale quantization model can perform reparameterization of each rotated residual in the rotated residual dataset into a direction component and a scale component. The multiscale quantization model can perform product quantization of the direction components of the plurality of rotated residuals, and perform scalar quantization of the scale components of the plurality of rotated residuals.

Abstract: An electronic device, for each media content item of a plurality of media content items, receives a respective score for each a first set of objectives and one or more other objectives and generates a respective score between a user and the media content item. The generating includes applying a first ordered weighted average to the respective scores for the first set of objectives, to produce a first combined score for the first set of objectives, applying a second ordered weighted average to the respective scores for a second set of objectives, wherein the second set of objectives includes (i) a resulting objective corresponding to the first set of objectives and having the first combined score and (ii) the one or more other objectives. The electronic device streams media content to the user selected based on the respective scores between the user and the media content items.

Abstract: Multidimensional digital content search techniques are described that support an ability of a computing device to perform search with increased granularity and flexibility over conventional techniques. In one example, a control is implemented by a computing device that defines a multidimensional (e.g., two-dimensional) continuous space. Locations in the multidimensional continuous space are usable to different search criteria through different weights applied to the criteria associated with the axes. Therefore, user interaction with this control may be used to define a location and corresponding coordinates that may act as weights to the search criteria in order to perform a search of digital content through use of a single user input.

Abstract: In an approach for a cooperative build and content annotation system for conversational design of virtual assistants, a processor formulates a build context based on a build activity of a user. A processor formulates one or more content queries based on the build context. A processor builds a content index by augmenting a text-search index with a neural Information Retrieval (IR) index. A processor searches the content index using the one or more content queries to identify content relevant to the build context. A processor determines at least one recommendation for the user based on heuristic rules applied to the build context and the identified content, wherein each recommendation is a build suggestion or a content annotation suggestion.

Abstract: An application instance that includes one or more machine learning models receives, from a subscriber computing system, a document comprising unstructured data. Based on the unstructured data, the application instance generates an optimized model input that includes a plurality of parsed document sections. For each parsed document section, the application instance generates an output set by performing, by a machine learning model, at least one key information extraction operation. The machine learning model transmits the output in structured form to a target application operated or hosted at least in part by a subscriber entity associated with the subscriber computing system.

Abstract: Systems and methods data structures queries to manage loading time of multimedia content are provided. A system receives an input audio signal from a computing device and identifies a request. The system identifies a quantized day value and a quantized time value. The system retrieves, via a lookup in a multi-dimensional matrix data structure with the quantized day value and quantized time value, signals for the quantized day value and the quantized time value. The system generates a query with the signals, and applies the query to a multimedia content data structure to identify a plurality of multimedia content items that match the plurality of signals. The system provides an indication of the plurality of multimedia content items that match the signals for the quantized day value and the quantized time value retrieved from the multi-dimensional matrix data structure.

Abstract: Computer-based systems and methods for risk/reward detection, visualization, and resolution using modular chainable algorithms are provided. The system allows for computer-based modeling of large data sets with improving processing speed and utilizing fewer computational resources. The modular chainable algorithms included embedded program code executable by a processor for performing a data modeling or analytic function on source data, visualization code for visualizing output of the program code, and workflow code for automatically performing one or more actions relating to the data modeling or analytic function.