HIERARCHICAL MULTI-ARMED BANDIT FOR LEXICON EXPANSION

    公开(公告)号:US20240184999A1

    公开(公告)日:2024-06-06

    申请号:US18062208

    申请日:2022-12-06

    摘要: Disclosed herein are methods, systems, and computer program products for selecting an artificial intelligence (AI) model. Aspects include receiving, by a multi-armed bandit agent, candidate entities from multiple set expansion models and selecting a subset of the candidate entities for a first set expansion task, wherein a first candidate entity of the selected subset of candidate entities is selected from a first model by a subject matter expert. Aspects also include selecting, by the multi-armed bandit agent, based on the first candidate entity selected by the subject matter expert, the first model from the set expansion models for generating further candidate entities.

    Index value giving apparatus, index value giving method and program

    公开(公告)号:US11960836B2

    公开(公告)日:2024-04-16

    申请号:US17442524

    申请日:2020-03-11

    摘要: An index value assignment device comprises: a selection unit that prompts a user to select, from a sentence, one or more words that the user feels to have a same degree of a predetermined index value with respect to one or more first words for which the index value is known; a validation unit that validates the selection result by the user on a basis of the index value of a second word for which the index value is known from among the one or more words included in the selection result; and a determination unit that determines the index value with respect to a third word for which the index value is unknown from among the one or more words selected by the user, on a basis of the index value of the first word according to the selection result for which the selection result is determined to be valid. Consequently, the efficiency of assigning word familiarity is increased.

    Unusual score generators for a neuro-linguistic behavioral recognition system

    公开(公告)号:US11914956B1

    公开(公告)日:2024-02-27

    申请号:US18088189

    申请日:2022-12-23

    IPC分类号: G06F40/237 G06F40/30

    CPC分类号: G06F40/237 G06F40/30

    摘要: Techniques are disclosed for generating anomaly scores for a neuro-linguistic model of input data obtained from one or more sources. According to one embodiment, generating anomaly scores includes receiving a stream of symbols generated from an ordered stream of normalized vectors generated from input data received from one or more sensor devices during a first time period. Upon receiving the stream of symbols, generating a set of words based on an occurrence of groups of symbols from the stream of symbols, determining a number of previous occurrences of a first word of the set of words, determining a number of previous occurrences of words of a same length as the first word, and determining a first anomaly score based on the number of previous occurrences of the first word and the number of previous occurrences of words of the same length as the first word.

    METHOD AND A SYSTEM FOR SUGGESTING AT LEAST ONE CAPTION FOR AN IMAGE

    公开(公告)号:US20240062565A1

    公开(公告)日:2024-02-22

    申请号:US18227720

    申请日:2023-07-28

    摘要: Provided is a method for suggesting a caption for an image, comprising: receiving the image; determining a plurality of impacting categories associated with a plurality of contextual keywords for the image, each of the plurality of impacting categories representing a sentiment associated with the plurality of contextual keywords; grouping the plurality of contextual keywords into a plurality of groups based on the plurality of impacting categories; determining an order associated with the plurality of contextual keywords, based on a pre-determined impacting function; generating at least one caption by processing each contextual keyword, based on the order associated with the plurality of contextual keywords; determining a priority value associated with each of the at least one caption based on information associated with the corresponding caption, a user profile, and the image; and suggesting the caption from the at least one caption based on the priority value.