SYSTEM AND METHOD FOR INDUCTIVE LEARNING ON GRAPHS WITH KNOWLEDGE FROM LANGUAGE MODELS

    公开(公告)号:WO2023072421A1

    公开(公告)日:2023-05-04

    申请号:PCT/EP2021/087326

    申请日:2021-12-22

    Abstract: A computer-implemented method for inductive learning on graphs is disclosed, wherein the graph (110) includes a plurality of entities (112), wherein relationships exist between the entities (112), and wherein the entities (112) and relationships have a name string. According to embodiments of the invention, the method comprises: (a) creating for each entity (112) of the graph (110), based on the entity's (112) name string, a related text corpus; (b) using a pretrained language model (150) to compute, from the related text corpus of an entity (112), a respective contextual entity embedding for each entity (112) of the graph (110); training a graph-based machine-learning, ML, model (210) by using, for each entity (112) of the graph (110), the computed entity embeddings; and repeating, for unseen entities, steps (a) and (b) and using the trained ML model (210) to perform inductive predictions for the unseen entities.

    CPU-AWARE INTELLIGENT RADIO CONTROLLER
    3.
    发明申请

    公开(公告)号:WO2023030695A1

    公开(公告)日:2023-03-09

    申请号:PCT/EP2022/052492

    申请日:2022-02-02

    Abstract: The present invention relates to a computer-implemented method of resource allocation for multiple virtualized base stations, vBSs (420), deployed in a shared radio computing platform (410), wherein the vBSs (420) share the same Radio Unit, RU (430), having a fixed number of Physical Resource Blocks, PRBs. With respect to an efficient utilization of both radio resources and available computing resources, the method comprises obtaining, by a radio controller, a set of metrics from the radio computing platform (410) for each of the vBSs (420) deployed; learning, by a learning agent (450) of the radio controller based on the obtained set of metrics, the traffic patterns and the CPU consumption of the vBSs (720) deployed; and computing, by the radio controller based on the knowledge of current CPU set configurations of the vBSs (420) and based on the learned traffic patterns and the CPU consumption of the vBSs (420), a CPU set configuration, L3 cache allocation and PRB assignment for each vBS (420) in such a way that at least one of the overall CPU overhead, the decoding time and SLA violations are minimized.

    FAST POST-QUANTUM CRYPTOGRAPHIC SORTITION
    4.
    发明申请

    公开(公告)号:WO2023016665A1

    公开(公告)日:2023-02-16

    申请号:PCT/EP2021/073178

    申请日:2021-08-20

    Abstract: The present invention relates to a computer-implemented method for execution of a cryptographic sortition among a group of parties (210, 220). According to an embodiment of the invention, the method comprises committing, by a first party (210) of the group, to a set of n party-specific secret keys k1,..., kn for a block cipher E; obtaining, by the first party (210) and at least a second party (220) of the group, a common input x and an index r; encrypting, by the first party (210), the input x with the r-th key kr of the committed keys k1,..., kn, thereby generating an output y1 of the block-cipher E, and publishing the output y1 together with the key kr used for encryption; and encrypting, by the second party (220), the common input x with the key kr published by the first party (210), thereby generating an output y1' of the block-cipher E, and comparing the generated output y1' with the output y1 published by the first party (210).

    ENTITY MATCHING WITH JOINT LEARNING OF BLOCKING AND MATCHING

    公开(公告)号:WO2023006224A1

    公开(公告)日:2023-02-02

    申请号:PCT/EP2021/071471

    申请日:2021-07-30

    Abstract: Embodiments of the invention provide a method of identifying entities from different data sources as matching entity pairs that refer to the same real-world object. With regard to a reduction of time and effort required to identify all possible matches from two heterogeneous data sources, the method comprises providing a set of labelling functions to determine the matching entities and the non-matching entities of a source data set (102) and a least one target data (104) set; selecting, from the provided set of labelling functions, a subset of labelling functions for training machine learning models for a blocking module (106) that aims at filtering out as many unmatched entity pairs as possible without missing any true matches and for a matching module (108) that aims at predicting matching results for the remaining entity pairs not filtered out by the blocking module (106); and jointly learning both a blocking model for the blocking module (106) and a matching model for the matching module (108) based on the available unlabeled entity pairs and the labelling functions of the selected subset of labelling functions.

    VORRICHTUNG ZUR BEFESTIGUNG AN EINEM BAUTEIL

    公开(公告)号:WO2023274461A1

    公开(公告)日:2023-01-05

    申请号:PCT/DE2022/200114

    申请日:2022-06-09

    Abstract: Eine Vorrichtung zur Befestigung an einem Bauteil mit einem Grundkörper (1) und einem Ausgleichselement (2), wobei das Ausgleichselement (2) mit dem Grundkörper (1) über ein Gewinde (14) verbunden ist und wobei das Ausgleichselement (2) einen Durchgang (4) mit einer Mitnahmeeinrichtung (5) für eine Schraube aufweist, ist dadurch gekennzeichnet, dass der Grundkörper (1) ein Grundelement (10) und einen formschlüssig mit dem Grundelement (10) verbundenen Haltering (9) aufweist.

    METHOD AND SYSTEM FOR PREDICTING AND CLASSIFYING EVENT SEQUENCES EMBEDDED IN A KNOWLEDGE GRAPH

    公开(公告)号:WO2022253449A1

    公开(公告)日:2022-12-08

    申请号:PCT/EP2021/066554

    申请日:2021-06-18

    Abstract: A computer-implemented method for event sequence forecasting of a process instance is disclosed. The method comprises building up and training a three- layered prediction model (160) including a first, a second and a third layer. The first layer is a graph embedding layer that assigns a fixed-dimensional graph embedding vector to each node and relation type in a fused event and knowledge graph (200) that contains available structural information including events, knowledge graph nodes, and the links between them. The second layer is an event embedding layer that assigns to each event of the process instance a fixed-dimensional event embedding vector. The third layer is a prediction layer that receives as input a sequence of event embeddings from the second layer and that generates as output a prediction of an unknown property of the event sequence used as input.

    METHOD AND SYSTEM FOR MONITORING OF A PHYSICAL ENVIRONMENT'S PRONENESS TO INFECTIOUS DISEASE TRANSMISSION

    公开(公告)号:WO2022242823A1

    公开(公告)日:2022-11-24

    申请号:PCT/EP2021/062977

    申请日:2021-05-17

    Abstract: A method for monitoring of a physical environment's proneness to infectious disease transmission, the method comprising: - in a training phase: obtaining unlabeled sensor data from sensors of the physical environment in order to provide a set of sensor features; generating, in particular by applying situation labeling functions, a labeling matrix that is fed to a generative model, wherein said generative model feeds a discriminative classifier model with probabilistic labels for the sensor features, wherein the probabilistic labels of the generative model are used for training the discriminative classifier model; determining, in particular by a feature selection optimizer entity, a subset of the sensor features based on an optimization procedure; and - in an operational phase: using, by the discriminative classifier model, the subset of sensor features for detecting predefined situations, which make the physical environment prone to infectious disease transmission. Furthermore, a corresponding system monitoring of a physical environment's proneness to infectious disease transmission is disclosed.

    VERFAHREN ZUM BESTIMMEN EINES SCHWINGUNGSVERHALTENS EINES ELEKTROMOTORS UND/ODER DESSEN EINBAUUMGEBUNG, SOWIE ENTSPRECHENDER ELEKTROMOTOR UND VENTILATOR

    公开(公告)号:WO2022233373A1

    公开(公告)日:2022-11-10

    申请号:PCT/DE2022/200076

    申请日:2022-04-27

    Applicant: ZIEHL-ABEGG SE

    Abstract: Es ist ein Verfahren zum Bestimmen eines Schwingungsverhaltens eines Elektromotors (1), insbesondere eines Elektromotors eines Ventilators, und/oder dessen Einbauumgebung offenbart, wobei eine Drehbewegung eines Rotors des Elektromotors in einem Bremsvorgang abbremsbar ist. Dabei umfasst das Verfahren: • Erzeugen eines Rucks durch Auslösen eines Bremsvorgangs oder eines Beschleunigungsvorgangs oder durch Ändern eines Bremsvorgangs oder eines Beschleunigungsvorgangs, • Erzeugen (S3) von erfassten Schwingungswerten durch Erfassen von Schwingungen zumindest eines Teils des Elektromotors mittels mindestens eines Schwingungssensors (9), • Bestimmen (S4) von spektralen Anteilen durch eine Frequenzanalyse der erfassten Schwingungswerte und • Bestimmen (S5) eines Schwingungsverhaltens des Elektromotors und/oder dessen Einbauumgebung durch Auswerten der spektralen Anteile. Des Weiteren ist ein entsprechender Elektromotor (1), ein Ventilator und ein System offenbart, die jeweils zum Ausführen des Verfahrens ausgebildet sein können.

    VENTILATOR, INSBESONDERE RADIAL- ODER DIAGONALVENTILATOR

    公开(公告)号:WO2022233372A1

    公开(公告)日:2022-11-10

    申请号:PCT/DE2022/200074

    申请日:2022-04-25

    Applicant: ZIEHL-ABEGG SE

    Abstract: Ventilator (1), insbesondere Radial- oder Diagonalventilator, mit einem Motor (4), einem vom Motor (4) drehangetriebenen Laufrad (3), einer Einlaufdüse (5) und einer sich um die Einlaufdüse (5) herum erstreckenden Düsenplatte (2), wobei das Laufrad (3) im Wesentlichen aus einer Bodenscheibe (10), einer Deckscheibe (8) und mehreren sich dazwischen erstreckenden Flügeln (9) besteht, die Düsenplatte (2) einen zur Druckseite hin umgekanteten Rand aufweist und die Deckscheibe (8) an ihrem Außenrand zur Saugseite hin abgerundet ist, und wobei die Umkantung (6) der Düsenplatte (2) und die Abrundung der Deckscheibe (8) derart geformt und dimensioniert sind, dass die deckscheibennahe Ausströmung aus dem Laufrad (3) mit der Abkantung der Düsenplatte (2) interagiert.

Patent Agency Ranking