International Business Machines Corporation

Andrew Gerald Crimmins

Pasquale A Catalano

John S Werner

Arkadiy O Tsfasman

Schayne Bellrose

Preston Lane

Anuradha Rao

Rorie Paul Reyes

Juan Merchan

Michael J Ellsworth, Jr

Prach Jerry Chuaypradit

Jerry Chuaypradit

Casimer M DeCusatis

Rajaram B Krishnamurthy

Christopher J Colonna

Samuel R Connor

Michael Onghena

Byron S Green

Faezeh Gholami

John Thomas Kinnear, Jr

Michael Lapointe

Robert K Overton

Mario L Borelli

Francesca Wisniewski

Nicholas W Crimmins

An example operation may include one or more of receiving primary data from one or more vehicles, extracting first information from the primary data, creating one or more documents based on the extracted information, generating one or more blockchain transactions based on the one or more documents, and committing the one or more documents to a blockchain.

Vehicle incident documentation for blockchain

Performing an unbiased product search. Keywords related to a product are received. A list of products is generated, based on product information in online content received in response to searches based on the keywords. An anonymized list of products is generated from the list of products by masking product brand names of the products in the list of products. For each product in the list of products, a list of online feedback items is generated. Based on credibility analysis and sentiment analysis of feedback items associated with a product, an aggregate rating score is assigned to the product. The list of products is ranked according to the aggregate rating scores. The ranked, anonymized list of products is presented on a user interface.

Unbiased search and user feedback analytics

A computer-implemented method includes identifying a first menu item on a standard menu of an establishment. A first plurality of ingredients of the first menu item are determined. A first subset of the first plurality of ingredients of the first menu item are categorized into a substitutable category. A second subset of the first plurality of ingredients of the first menu item are categorized into a non-substitutable category. Each ingredient of the first menu item that is in the substitutable category is compared to user preferences. A substitutable ingredient of the substitutable category is excluded, based on the comparisons, where the substitutable ingredient is in the first menu item. An available substitute is automatically identified for the substitutable ingredient of the substitutable category. A personalized menu of menu items selected from the standard menu is generated, where the personalized menu includes the first menu item modified with the available substitute.

Generating a personalized menu for submitting a custom order

A method and system for digital video content mapping includes receiving and parsing digital video content. Content tokens are defined for tagging content in the digital video content. A cognitive analysis is used to identify content tokens and tag content tokens in the digital video content. A relationship map is created between related tagged content tokens between instances of digital video content using a cognitive analysis. The relationship map indicates the strength of a relationship between the tagged plurality of content tokens and thereby the plurality of instances of digital video content.

Video content relationship mapping

Anonymizing product search results. Keywords related to a product are received. For each keyword, an associated list of terms, which lists the keyword and one or more additional terms that are closely related to the keyword, is generated. A combination list that includes combinations of terms selected from the associated lists of terms, where each combination includes two or more terms and at least one combination includes terms from different associated lists of terms, is created. A search, based on the combinations of terms in the combination list, is performed and online content is received. Based on product information in the online content, a list of products is generated. An anonymized list of products is generated from the list of products by masking product brand names of the products in the list of products. The anonymized list of products is presented on a user interface.

Unbiasing search results

A computer-implemented method may include obtaining, by processor, a first message composed in a first language and obtaining, by the processor, a translated first message. The translated first message may include a translation of the first message from the first language to a second language. The method may further include determining, by the processor, that the translated first message includes a translation-generated additional meaning, and notifying, by the processor, the first user of the determination.

Addressing additional meanings resulting from language translation

A computer-implemented method includes initializing a navigation application on a device, determining a current location of the device, and receiving a destination. The method also includes determining whether a known geographical feature is located on a travel path to the destination. The known geographical feature may be derived from a user profile. The method, in response to determining that a known geographical feature is not located on the travel path to the destination, includes outputting graphical and audio directions to the destination. In response to determining that a known geographical feature is located on the travel path to the destination, a sub-method includes determining whether the device is currently in a known area and outputting graphical and audio directions to the known geographical feature in response to determining that the device is not currently in the known area.

Selectively generating directions based on device location

Aspects include obtaining, by a sending system, a measured receive optical power level of an optical signal that was received at a receiving system coupled to the sending system via an optical network. The optical signal was sent via an optical transmitter of the sending system to an optical receiver of the receiving system. An optimal receive optical power level of the optical receiver of the receiving system is determined by the sending system. The sending system adjusts an output optical power level of the optical transmitter in response to determining that the measured receive optical power level is not within a threshold of the optimal receive optical power level. The adjusting is performed without decoupling the sending system from the receiving system.

Transmit optical power adjustment based on receive optical power measurements

A cognitive hive architecture for determining and managing a hive health index of a hive. The cognitive hive architecture is executable by a processor in communication with a computing device or a computing environment. The cognitive hive architecture includes sensors embedded within the hive. The cognitive hive architecture acquires data from the sensors, which are monitoring real-time conditions of the hive. The data being representative of the real-time conditions. The cognitive hive architecture tags the data to physical identifications and positional identifications to produce tagged data. The cognitive hive architecture associates the tagged data to physical profiles and positional profiles and generates the hive health index based on the physical profiles and the positional profiles.

Cognitive hive architecture

A processor implemented method for evaluating a written communication in view of a proficiency of at least one recipient of the written communication is provided. The processor implemented method includes determining a proficiency profile of at least one user with respect to at least one knowledge area based on data sources read or created by the at least one user. The processor implemented method includes evaluating the written communication by identifying a proficiency of the written communication, comparing the proficiency of the written communication to the proficiency of the at least one recipient, and providing feedback to a creator of the written communication as to whether the at least one recipient understands the written communication.

Predicting if a message will be understood by recipients

A method for improving traffic flow at an intersection includes determining spacing data for a plurality of vehicles within traffic lanes that enable the plurality of vehicles to proceed to an intersection from a common direction, determining whether traffic flow proximate to the intersection is hindered due to spacings between the plurality of vehicles, and sending a request to at least one vehicle of the plurality of vehicles to move forward to improve the traffic flow proximate to the intersection. The spacing data may be derived from spacing-related sensor data captured by one or more vehicles of the plurality of vehicles. A corresponding apparatus and computer program product for executing the above method are also disclosed herein.

Traffic flow at intersections

Embodiments include method, systems and computer program products for addressing packet loss over a Voice over Internet Protocol (VoIP) network using phonemic restoration. The method includes receiving, by a processor, a plurality of data packets associated with a VoIP call. The processor further determines that one or more of the plurality of data packets have been lost or discarded. The processor further inserts, a sound into the VoIP call in response to the one or more of the plurality of data packets being lost or discarded. The processor further sends the sound to an output device in order to trigger a phonemic restoration response by a listener.

Addressing packet loss in a voice over internet protocol network using phonemic restoration

Examples of techniques for predictive media content delivery are disclosed. In one example implementation according to aspects of the present disclosure, a computer-implemented method for predictive media content delivery includes identifying, by the processing device, a customer approaching a kiosk. The method further includes determining, by the processing device, an estimated duration that the customer is expected to be in proximity to the kiosk. The method further includes presenting, by the processing device, media content to the customer that has a playback duration commensurate with the estimated duration that the customer is expected to be in proximity to the kiosk.

Predictive media content delivery

Embodiments include techniques for determining the validity of service recommendations, where the techniques include receiving a service provider recommendation for a device from a service provider, and receiving device inputs and service provider inputs associated with the device. The techniques also include performing an input analysis on the device inputs to determine a predicted recommendation, and determining a trust level score for the service provider based at least in part on the service provider inputs, and comparing the service provider recommendation and the predicted recommendation. Techniques include performing, based at least in part on the trust level score, a value analysis and a severity analysis, and generating a recommended action based at least in part on the value analysis and the severity analysis.

Growing community of inventors

Montrose, NY, United States of America

Andrew Gerald Crimmins