Abstract: This application relates generally to intelligent and explainable link prediction in knowledge graph systems that automatically incorporate user feedback. In one aspect, this application discloses an iterative process for predicting a link set as a group of links in a knowledge graph in an embedding space by expanding the knowledge graph with predicted and validated single links in each iteration such that a final set of links are predicted with each one being added to the set depending on previously added predicted links. In another aspect, this application also discloses automatically extracting rules from user feedback of link predictions and generating a user feedback knowledge graph from the extracted rules, which in combination with an original knowledge graph are used for the generation of the link predictions.

Abstract: This application relates generally to intelligent and explainable link prediction in knowledge graph systems that automatically incorporate user feedback. In one aspect, this application discloses an iterative process for predicting a link set as a group of links in a knowledge graph in an embedding space by expanding the knowledge graph with predicted and validated single links in each iteration such that a final set of links are predicted with each one being added to the set depending on previously added predicted links. In another aspect, this application also discloses automatically extracting rules from user feedback of link predictions and generating a user feedback knowledge graph from the extracted rules, which in combination with an original knowledge graph are used for the generation of the link predictions.

Abstract: Methods and systems for estimating causal effects from knowledge graphs are provided. The method includes obtaining intervention application data and subject history data for a candidate group of subjects, and dividing, based on the intervention application data, the candidate group into a reception subgroup that received an intervention and a rejected subgroup that did not receive the intervention. The method includes for each subject within the candidate group, mapping, based on the subject history data, a covariate value set onto a knowledge graph with an embedding neural network; and for each subject in the reception subgroup or the rejection subgroup, translating the covariate value sets for the subjects within the reception subgroup or the rejection subgroup into a reception matrix or a rejection matrix with a feature neural network. The method includes comparing the reception subgroup to the rejection subgroup to determine a differential intervention effect.

Abstract: Complex computer system architectures are described for analyzing data elements of a knowledge graph, and predicting new facts from relational learning applied to the knowledge graph. This discovery process includes converting the knowledge graph into a set of candidate embeddings spaces to apply further analysis to rank the set of candidate embeddings spaces, where the top ranked candidate embeddings spaces are further processed to identify the new facts.

Abstract: A device may receive data associated with units of a group. A subset of the data, for a unit, may include values for features of the unit and an indication of whether the subset indicates that the unit satisfies a qualification threshold of a qualification model. The device may identify subsets of the data that indicate that a subsets of units do not satisfy the qualification threshold; alter feature values, of the subsets of the units, for a feature to generate revised subsets of the data; and process, based on the qualification model, the revised subsets of the data to obtain counterfactual explanations. The device may determine an impact score associated with the feature based on a quantity of units, of the subset of units, that satisfied the qualification threshold based on the revised subsets of the data; and determine that the impact score satisfies an impact threshold.

Abstract: A device may receive calibration data associated with a plurality of units and receive a set of rules; determine, based on the set of rules, a plurality of groups associated with the plurality of units; and process the calibration data based on a pretrained artificial intelligence (AI) model. The device may determine, based on processing the calibration data, a prediction that is associated with a group of the plurality of groups; and determine, based on the set of rules, a target associated with the group based on the set of rules. The device may generate a calibration model based on the prediction and the target, and aggregate the calibration model with another calibration model that is associated with another group of the plurality of groups to form a calibrated AI model.

Abstract: Complex computer system architectures are described for analyzing data elements of a knowledge graph, and predicting new facts from relational learning applied to the knowledge graph. This discovery process includes converting the knowledge graph into a set of candidate embeddings spaces to apply further analysis to rank the set of candidate embeddings spaces, where the top ranked candidate embeddings spaces are further processed to identify the new facts.

Abstract: A method to mitigate the consequences of a vapor cloud explosion due to an accidental release of a flammable gas in an open area, comprising: defining an hazardous area wherein an accidental release of flammable gas is likely to happen; receiving a signal from a detector device able to detect the presence of the flammable gas within the hazardous area, upon reception of a signal indicating the presence of the flammable gas within the hazardous area, generating a control signal to activate a release of a flame acceleration suppression product in the hazardous area, at a rate that is determined as a function of the volume of said hazardous area.

Abstract: A method to mitigate the consequences of a vapor cloud explosion due to an accidental release of a flammable gas in an open area, comprising: defining an hazardous area wherein an accidental release of flammable gas is likely to happen; receiving a signal from a detector device able to detect the presence of the flammable gas within the hazardous area, upon reception of a signal indicating the presence of the flammable gas within the hazardous area, generating a control signal to activate a release of a flame acceleration suppression product in the hazardous area, at a rate that is determined as a function of the volume of said hazardous area.

Abstract: The invention relates to a method for treating a hydrocarbons charge comprising the following stages, in which: a) the charge is brought into contact with a solvent in order to obtain a deasphalted effluent having a content of asphaltenes below 3000 ppm by weight, b) the deasphalted effluent is cracked in the presence of hydrogen and a hydrocracking catalyst, in a bubbling-bed reactor, so as to convert at least 50 wt. % of the fraction of the deasphalted effluent boiling above 500° C. to compounds having a boiling point below 500° C., c) the effluent from stage b) is fractionated to recover gasolines, kerosene, gas oils and a first residue, and d) at least a portion of this first residue is cracked so as to obtain an effluent comprising gasolines, kerosene, gas oils and a second residue.

Abstract: The present invention concerns an improved once-through process for hydrocracking hydrocarbon-containing feeds with high nitrogen contents, with partial elimination of ammonia, for example by a hot flash located between the hydrorefining zone and the hydrocracking zone. Said hot flash drum functions at a pressure close to that of the hydrorefining reactor and at a temperature in the range from 150° C. to the hydrorefining reactor outlet temperature.

Abstract: The invention concerns a process for producing middle distillates from a paraffinic feed produced by Fischer-Tropsch synthesis, using a hydrocracking/hydroisomerization catalyst which comprises 0.2% to 2.5% by weight of an oxide of a doping element selected from boron, phosphorus, silicon, at least one hydrodehydrogenating element selected from the group formed by noble elements from group VIII of the periodic table, a non-zeolitic support based on silica-alumina containing more than 5% by weight and 95% by weight or less of silica (SiO2), specifically defined pore characteristics, a BET specific surface area in the range 100 to 550 m2/g, and with an X ray diffraction diagram which contains at least the characteristic principal peaks of at least one of the transition aluminas included in the group composed of alpha, rho, khi, eta, gamma, kappa, theta and delta aluminas.

Abstract: The present invention concerns doped catalysts on a mixed zeolite/alumino-silicate support with a low macropore content, and hydrocracking/hydroconversion and hydrotreatment processes employing them. The catalyst comprises at least one hydrodehydrogenating element selected from the group formed by elements from group VIB and group VIII of the periodic table and a doping element in a controlled quantity selected from phosphorus, boron and silicon, and a support based on zeolite Y defined by a lattice parameter a of the unit cell in the range 24.40×10?10 m to 24.15×10?10 m and silica-alumina containing a quantity of more than 5% by weight and 95% by weight or less of silica (SiO2).

Abstract: The invention relates to a method for treating a hydrocarbons charge comprising the following stages, in which: a) the charge is brought into contact with a solvent in order to obtain a deasphalted effluent having a content of asphaltenes below 3000 ppm by weight, b) the deasphalted effluent is cracked in the presence of hydrogen and a hydrocracking catalyst, in a bubbling-bed reactor, so as to convert at least 50 wt. % of the fraction of the deasphalted effluent boiling above 500° C. to compounds having a boiling point below 500° C., c) the effluent from stage b) is fractionated to recover gasolines, kerosene, gas oils and a first residue, and d) at least a portion of this first residue is cracked so as to obtain an effluent comprising gasolines, kerosene, gas oils and a second residue.

Abstract: The invention relates to a catalyst including at least one hydro-dehydrogenating element chosen from the group formed by the group VIB and group VIII elements of the periodic table and a substrate based on a silica-alumina matrix with a reduced content of macropores containing a quantity greater than 5% by weight and less than or equal to 95% by weight of silica (SiO2) and based on at least one zeolite. The invention also relates to a substrate based on a silica-alumina matrix with a reduced content of macropores containing a quantity greater than 5% by weight and less than or equal to 95% by weight of silica (SiO2) and based on at least one zeolite. The invention also relates to hydrocracking and/or hydroconversion processes and hydrotreating processes utilizing a catalyst according to the invention.

Abstract: The invention relates to a process for FCC pretreatment by mild hydrocracking of a hydrocarbon feedstock that comprises a vacuum distillate fraction or a deasphalted oil or else a mixture of these two fractions, said primary feedstock, to produce gas oil and an effluent having an initial boiling point of more than 320° C., said effluent (FCC feedstock) then being subjected to a catalytic cracking, process in which at least 85% by weight of said primary feedstock ends above 375° C. and at least 95% by weight of said primary feedstock ends below 650° C., whereby the mild hydrocracking is performed under an absolute pressure of 2 to 12 MPa and at a temperature of between 300 and 500° C., characterized in that the hydrocarbon feedstock also comprises a lighter hydrocarbon fraction, a so-called secondary feedstock, of which at least 50% by weight ends below 375° C. and at least 80% ends above 200° C.

Abstract: Process for FCC pretreatment by mild hydrocracking including diluting the feedstock with a feedstock of biological origin The invention relates to a process which comprises the stages consisting in carrying out a mild hydrocracking, implemented under a pressure of 20 to 120 bar and at a temperature of between 300 and 500° C., on a hydrocarbon feedstock comprising or consisting of a feedstock of petroleum origin, comprising or consisting of a vacuum distillate fraction and/or a deasphalted oil, at least 85% by weight of which boils above 370° C., and a feedstock of biological origin comprising components of vegetable oil and/or animal fat type, in order to produce gas oil and an effluent having an initial boiling point of greater than 330° C. within the meaning of the simulated distillation according to standard ASTM D2287, and carrying out a catalytic cracking (FCC) of a composition comprising or consisting of said effluent, and to a plant which makes it possible to carry out this process.

Abstract: The invention concerns a process for producing middle distillates from a paraffinic feed produced by Fischer-Tropsch synthesis, using a hydrocracking/hydroisomerization catalyst which comprises: at least one hydrodehydrogenating element selected from the group formed by elements from group VIB and group VIII of the periodic table; 0.01% to 6% of phosphorus as a doping element; and a non-zeolitic support based on mesoporous alumina-silica.

Abstract: The invention relates to a process, including removal of resins, for the treatment of a hydrocarbon charge, at least 80% of the compounds of which have a boiling point which is above or equal to 340° C., in which process: the charge is sent to a fractionation stage during which the recovery takes place of at least one heavy fraction and at least one light fraction, at least some of the heavy fraction is sent to an extraction stage during which resins contained in said heavy fraction are extracted, and a purified fraction is recovered, a mixture is made which comprises at least part of the purified fraction which was obtained in the extraction stage and at least one light fraction which was obtained in the fractionation stage, and the mixture thus obtained is sent to a cracking stage.

Abstract: This invention relates to doped catalysts on an aluminosilicate substrate with a low content of macropores and the hydrocracking/hydroconversion and hydrotreatment processes that use them. The catalyst comprises at least one hydro-dehydrogenating element that is selected from the group that is formed by the elements of group VIB and group VIII of the periodic table and a dopant in a controlled quantity that is selected from among phosphorus, boron, and silicon and a non-zeolitic substrate with a silica-alumina base that contains a quantity of more than 15% by weight and of less than or equal to 95% by weight of silica (SiO2).