Abstract: One embodiment provides a method that includes obtaining a default language corpus. A second language corpus is obtained based on a second language preference. A first transcription of an utterance is received using the default language corpus and natural language processing (NLP). At least one problem word in the first transcription is determined based on an associated grammatical relevance to neighboring words in the first transcription. Upon determining that a first probability score is below a first threshold, an acoustic lookup is performed for an audible match for the problem word in the first transcription based on an associated acoustical relevance. Upon determining that a second probability score is below a second threshold, it is determined whether a match for the problem word exists in the secondary language corpus. Upon determining that the match exists in the secondary language corpus, a second transcription for the utterance is provided.

Abstract: One embodiment provides a method that includes obtaining a default language corpus. A second language corpus is obtained based on a second language preference. A first transcription of an utterance is received using the default language corpus and natural language processing (NLP). At least one problem word in the first transcription is determined based on an associated grammatical relevance to neighboring words in the first transcription. Upon determining that a first probability score is below a first threshold, an acoustic lookup is performed for an audible match for the problem word in the first transcription based on an associated acoustical relevance. Upon determining that a second probability score is below a second threshold, it is determined whether a match for the problem word exists in the secondary language corpus. Upon determining that the match exists in the secondary language corpus, a second transcription for the utterance is provided.

Abstract: One embodiment provides a method for predicting a next action in a conversation system that includes obtaining, by a processor, information from conversation logs and a conversation design. The processor further creates a dialog graph based on the conversation design. Weights and attributes for edges in the dialog graph are determined based on the information from the conversation logs and adding user input and external context information to an edge attributes set. An unrecognized user input is analyzed and a next action is predicted based on dialog nodes in the dialog graph and historical paths. A guiding conversation response is generated based on the predicted next action.

Abstract: In an aspect of the present disclosure, a method for providing an alternate modality of input for filling a form field in response to a failure of voice recognition is disclosed including prompting the user for information corresponding to a field of a form, generating speech data by capturing a spoken response of the user to the prompt using at least one input device, attempting to convert the speech data to text, determining that the attempted conversion has failed, evaluating the failure using at least one speech rule, selecting, based on the evaluation, an alternate input modality to be used for receiving the information corresponding to the field of the form, receiving the information corresponding to the field of the form from the alternate input modality, and injecting the received information into the field of the form.

Abstract: One embodiment provides a method comprising generating a conversational interface for display on an electronic device. The conversational interface facilitates a communication session between a user and an automated conversational agent. The method further comprises performing a real-time analysis of a portion of a user input in response to the user constructing the user input during the communication session, and updating the conversational interface to include real-time feedback indicative of whether the automated conversational agent understands the portion of the user input based on the analysis. The real-time feedback allows the user to adjust the user input before completing the user input.

Abstract: One embodiment provides a method comprising maintaining a weather model based on predicted weather conditions for an air traffic control zone. A hash table comprising multiple hash entries is maintained. Each hash entry comprises a timestamped predicted weather condition for a cell in the zone. A flight plan request for a drone is received. The request comprises a planned flight path for the drone. For at least one cell on the planned flight path, same latitude or same longitude cells, whichever is most closely orthogonal to a direction of the planned flight path, are heuristically probed. Weather conditions for the at least one cell are estimated based on predicted weather conditions for the same latitude or same longitude cells. An executable flight plan is generated if the planned flight path is feasible based on the estimated weather conditions; otherwise, a report including an explanation of infeasibility is generated instead.

Abstract: Embodiments of the present invention provide an apparatus comprising a body including a cavity for storing one or more packages, and a conveyor belt disposed above a top surface of the body. The belt is shaped to receive one or more packages, and the belt is controllable to rotate a package placed on the belt either from the top surface to the cavity for storage or from the cavity to the top surface for dispatch. A package comprises at least one of a drone and a payload transported by the drone. The apparatus further comprises a landing mechanism for stabilizing a drone landing on the apparatus.

Abstract: One embodiment provides a method for predicting a next action in a conversation system that includes obtaining, by a processor, information from conversation logs and a conversation design. The processor further creates a dialog graph based on the conversation design. Weights and attributes for edges in the dialog graph are determined based on the information from the conversation logs. An unrecognized user input the conversation system is detected. The unrecognized user input is analyzed and a next action is predicted based on dialog nodes in the dialog graph and historical paths. A guiding conversation response is generated based on the predicted next action.

Abstract: One embodiment provides a method including receiving authoring conversational training data. A machine learning based conversational agent is trained with the conversational training data. The training includes: creating and storing example transcripts of user utterances, creating and storing example transcripts of agent utterances, sequencing utterance transcripts using the example transcripts of user utterances and the example transcripts of agent utterances, forming a corpus from the sequenced utterance transcripts, marking speech patterns that represent social actions from tagging the sequenced utterance transcripts, and forming a patterned corpus from the marked speech patterns.

Abstract: One embodiment provides a computer program product for improving accuracy of a transcript of a spoken interaction. The computer program product comprises a computer readable storage medium having program instructions embodied therewith. The program instructions are executable by a processor to cause the processor to identify a plurality of patterns in the transcript. The plurality of patterns are indicative of a group of acoustically similar words in the transcript and a corresponding local, sequential context of the group of acoustically similar words. The program instructions are further executable by the processor to cause the processor to predict conditional probabilities for the group of acoustically similar words based on a predictive model and the plurality of patterns, detect one or more transcription errors in the transcript based on the conditional probabilities, and correct the one or more transcription errors by applying a multi-pass correction on the one or more transcription errors.

Abstract: One embodiment provides a method comprising intercepting an incoming communication for a user, and determining whether to classify the incoming communication as a suspicious incoming communication based on content of the incoming communication and a learned classification model or learned signatures. The method further comprises monitoring one or more outgoing communications from the user, and invoking an action relating to scam prevention in response to determining an outgoing communication from the user is linked to a suspicious incoming communication.

Abstract: One embodiment provides a method comprising intercepting a voice communication, collecting multi-sensory inputs associated with the voice communication, and determining an overall risk assessment metric for the voice communication based on the multi-sensory inputs and learned signatures. The multi-sensory inputs are indicative of content of the voice communication and one or more contextual factors associated with a target of the voice communication. The overall risk assessment metric indicates a likelihood the voice communication is a scam.

Abstract: Technical solutions are described for analyzing a natural language conversation-generating machine. The computer implemented method includes identifying, from a conversation log, a plurality of adjacency pairs. The method further includes determining, from the adjacency pairs, a number of adjacency pairs with outcome success indicators. The method further includes determining, from the adjacency pairs, a number of adjacency pairs with outcome failure indicators. The method further includes computing a mutual understanding score for the machine by computing a ratio of the number of adjacency pairs with outcome success indicators and the number of adjacency pairs with outcome failure indicators.

Abstract: One embodiment provides a method comprising answering one or more incoming phone calls received at one or more pre-specified phone numbers utilizing a bot. The bot is configured to engage in a conversation with a caller initiating an incoming phone call utilizing a voice recording that impersonates a human being. The method further comprises recording each conversation the bot engages in, and classifying each recorded conversation as one of poison data or truthful training data based on content of the recorded conversation and one or more learned detection models for detecting poisoned data.

Abstract: One embodiment provides a method comprising intercepting an incoming communication for a user, and determining whether to classify the incoming communication as a suspicious incoming communication based on content of the incoming communication and a learned classification model or learned signatures. The method further comprises monitoring one or more outgoing communications from the user, and invoking an action relating to scam prevention in response to determining an outgoing communication from the user is linked to a suspicious incoming communication.

Abstract: A computer-implemented method according to one embodiment includes determining a starting folder within a file system, computing, for each child folder of the starting folder, a similarity metric indicating a level of similarity to a file, selecting two child folders of the starting folder having greatest similarity metrics, comparing a difference between the greatest similarity metrics of the two child folders to a predetermined threshold, and conditionally selecting the starting folder as a recommended folder to which the file is saved, based on the comparing.

Abstract: One embodiment provides a computer program product for improving accuracy of a transcript of a spoken interaction. The computer program product comprises a computer readable storage medium having program instructions embodied therewith. The program instructions are executable by a processor to cause the processor to identify a plurality of patterns in the transcript. The plurality of patterns are indicative of a group of acoustically similar words in the transcript and a corresponding local, sequential context of the group of acoustically similar words. The program instructions are further executable by the processor to cause the processor to predict conditional probabilities for the group of acoustically similar words based on a predictive model and the plurality of patterns, detect one or more transcription errors in the transcript based on the conditional probabilities, and correct the one or more transcription errors by applying a multi-pass correction on the one or more transcription errors.

Abstract: One embodiment provides a method comprising answering one or more incoming phone calls received at one or more pre-specified phone numbers utilizing a bot. The bot is configured to engage in a conversation with a caller initiating an incoming phone call utilizing a voice recording that impersonates a human being. The method further comprises recording each conversation the bot engages in, and classifying each recorded conversation as one of poison data or truthful training data based on content of the recorded conversation and one or more learned detection models for detecting poisoned data.

Abstract: According to one embodiment, a computer-implemented method for dynamically modifying placeholder text in a conversational interface includes: processing a conversation log reflecting a conversation between a human user and an automated agent; determining, based at least in part on the processing: one or more capabilities of the automated agent; and/or a trajectory of the conversation; and dynamically modifying placeholder text in the conversational interface based at least in part on: the one or more capabilities of the automated agent; the trajectory of the conversation; or both the one or more capabilities of the automated agent and the trajectory of the conversation. Other embodiments in the form of systems and computer program products are also disclosed.

Abstract: One embodiment provides a method comprising generating a conversational interface for display on an electronic device. The conversational interface facilitates a communication session between a user and an automated conversational agent. The method further comprises performing a real-time analysis of a portion of a user input in response to the user constructing the user input during the communication session, and updating the conversational interface to include real-time feedback indicative of whether the automated conversational agent understands the portion of the user input based on the analysis. The real-time feedback allows the user to adjust the user input before completing the user input.