Digital Commons Network^™

Efficient E-Commerce On Virtual Personal Assistants, Thomas Price, Justin Lewis

Defensive Publications Series

Virtual personal assistants perform for their users a variety of tasks, e.g., placing online shopping orders, performing e-commerce transactions, handle shopping and commerce queries, etc. Typically, virtual personal assistants interact conversationally with their users and adapt to their user’s preferences using machine intelligence. The conversational model between users and virtual assistants makes it difficult for users to get product details needed to confidently place an online order. It is also difficult, through conversation alone, to input delivery information, authenticate payment, or even be confident of what is being purchased. Techniques of this disclosure facilitate e-commerce over virtual personal assistants by …

Placing Sponsored-Content Associated With An Image, Yushi Jing, Shumeet Baluja

Defensive Publications Series

Techniques are described for placing sponsored-content associated with an image. The techniques may include matching a first image for which a sponsored-content item is to be selected with a reference image. A sponsored-content item to be presented may be selected based on an association between the reference image and the sponsored-content item to be presented.

Sub-Type Checking At Runtime With Dynamic Class Loading, Igor Murashkin, Mingyao Yang

Defensive Publications Series

Subtype checking is common in computer programs to determine whether an object or an array is a subtype of a class. This check is typically performed at runtime. Traditional techniques for subtype checking are slow since such techniques often need to traverse an entire class hierarchy to return a result. The techniques of this disclosure offer a faster alternative. Each class in the hierarchy is encoded. Leading bits or prefix in the encoding are used to determine, during dynamic class loading, whether a subclass relationship exists between a source class and a target class.

Smart Attenuation Control For Microphones, Sara Basson, Dimitri Kanevsky, Yiteng Huang

Defensive Publications Series

Microphone arrays are used to home into a source of sound. They find application in a variety of situations, e.g., conference rooms, video conferencing, virtual personal assistants, etc. When multiple speakers are participating in a joint conversation, microphone arrays tend to direct themselves to the loudest or clearest speakers. The problem with doing so is that the array does not provide timely and sufficient access to speakers with soft voices, accented speech, speech disorders, or whose speech is otherwise difficult to comprehend, e.g., low or unintelligible speech level. This disclosure describes techniques to identify people whose speech intelligibility levels are …

Zero-Javascript Stateful User Interface Framework, Matthew Steven Frisbie

Defensive Publications Series

To create a web page with state, a developer may not be required to use JavaScript. The developer may be able to use HTML and CSS to create a lightweight web page that can remember whether a user clicked on a web page or the location of the user click or selection without JavaScript. The developer may use checkboxes or radio buttons or both to create web pages with state without using JavaScript.

A Framework For Large-Scale Fault-Tolerant Data Pipelines, Long Nguyen, Zachary Frazier

Defensive Publications Series

A system and a method for frequent and fault-tolerant running of large-scale data pipelines across data centers are disclosed. The system includes one or more data centers, each having multiple servers, with a framework including a mutex design pattern for multi-binary data processing. The mutex design includes a minimal locking scheme where locks are acquired on the outputs of the stages on startup from a distributed locking service, and are released upon failure or successful writing of the output. A multi-homed pipeline may be utilized to make the system fault-tolerant. The method wraps each binary in the pipeline with locking …

Generation Of Speech Training Data For Special Speech Recognition Tasks, Dimitri Kanevsky, Andrew Senior, Sara Basson

Defensive Publications Series

Speech recognition is widely used as a voice-user interface in several settings, e.g., interactive voice response, virtual personal assistants, transcription, translation applications, etc. Although speech recognition technology has advanced far enough to be useful to a sizeable number of human speakers, there are still populations that cannot take full advantage of speech recognition. For example, people with impaired speech, speakers of rare languages or dialects, with strong accents, etc. have difficulty using an application that uses speech recognition. The reason for such user difficulty is that there is insufficient data to train an automatic speech recognizer to recognize such relatively …

Promoting High Quality Ad Data From Less-Prominent Formats To More-Prominent Representation, Ashwin Kumar, Sanchit Gupta

Defensive Publications Series

A system and method are disclosed that extracts high quality data from less-prominent text format to a more-prominent text format. The system extracts high quality advertiser-provided data found in low-prominent text formats such as callouts and content within site links. The extracted information is combined with available ad content present in user-noticeable text formats that are highly prominent. The information may include price info (in price extension), shipping info, service availability, return/pickup information, benefits, etc. When the user makes a commercial query, the system presents the combined ad content in the prominent text formats. The disclosed system increases the visibility …

Increased Efficiency In Ad Modeling Through Adveritser Analytic Power-User Analysis, Alex Jacobson

Defensive Publications Series

Increased efficiency in modeling through analytic power user analysis is introduced in this publication. The analytic power user analysis includes identifying a first set of users from a global set of users. The first set of users is generally characterized as proficient, relative to the global set, in setting up and/or tracking a content item. Interactions of the first set of users are analyzed to determine interactions or information associated with interactions that distinguish the first set of users from the global set. A second set of users are identified based on the distinguishing interactions or associated information. Performance data …

Optimized Route Options Based On Weather Conditions, Dina Betser

Defensive Publications Series

A weather based routing system is used to provide a routing and navigation option based on specific weather conditions and time parameters. The system receives a request to generate a route between a start point and an end point based on a weather condition for a specified time. These weather conditions can indicate picking the shadiest route, or the sunniest route, or the route which would be least windy between the start and the end point. The system can access a database of stored photographic maps data to get information about the two points and the different routes between them. …

Construction Of A Temporal Tree And Its Coarse Grain View In A Streaming Big Data Context, Crina Cristiana Arsenie, Christophe Girardin

Defensive Publications Series

Structuring data as a tree can apply in every day’s life (like friends, family network of connections etc.), but also in industries like Telecommunications. With the fast pace of technology advancements, these structures are growing, changing, bringing therefore a dynamicity that has to be managed. This disclosure provides an efficient technique to construct trees in a Streaming Big Data context and to optimize the processing of these tree updates in real time by other consumer applications.

Method For Preventing Excessive Trimming Of A Previously Written Data Track When Appending Data In A Magnetic Tape Recording Device, Jeffrey S. Mcallister

Defensive Publications Series

This invention provides a means for protecting previously written data tracks on magnetic tape from excessive trimming due to tape expansion or contraction caused by changes in environmental conditions. The width of tape media varies based on environmental conditions. If data tracks are written on a magnetic tape at one environmental condition and then data is later written to the same tape at a significantly different environmental condition, the previously written tracks can be over written to the point that the data on the previously written tracks can no longer be read. This invention measures the width of the data …

Using The Least Congested Path For Non-Minimal Load-Balanced Routing (Lcp), Nicholas George Mcdonald

Defensive Publications Series

The most popular non-minimal adaptive routing algorithms (namely UGAL) decide at the source router whether to route minimally to the destination or to route minimally to a randomly chosen intermediate node then minimally to the destination. The random intermediate selection implements Valiant's randomized routing which provides perfect load balancing, albeit at a reduced injection rate. In contrast, this methodology chooses the intermediate node to be a router that lies on the least congested output path of the current router.

Skipping Dimensions In Dimension Order Routing Algorithms (Sdor), Nicholas George Mcdonald, Michael Isaev

Defensive Publications Series

Dimension order routing algorithms use the least amount of network resources compared to other routing algorithms. Dimension order algorithms have lower amount of path diversity compared to other routing algorithms. This methodology establishes a methodology to allow dimension order algorithms to increase their path diversity by increasing the amount of resources required. This methodology implements dimensional skipping where the routing algorithm is able to determine that a particular dimensional plane is congested and that it should be skipped and revisited later on a different plane.

Obfuscator Overlay For Fingerprint Sensor, David Ogutu

Defensive Publications Series

Fingerprint authentication is a convenient technique for authentication on consumer devices such as smartphones, tablets, laptop computers, etc. However, some users have concerns regarding privacy of their fingerprint data. This disclosure addresses such privacy concerns while retaining the ability for users to utilize fingerprint authentication. Per this disclosure, an obfuscating capacitive layer along with a capacitive sensor is placed on a fingerprint sensor, or is integrated in a fingerprint sensor. The layer generates a random pattern at the time of creation of the layer. This pattern along with the user’s fingerprint is recognized and used to authenticate a user. In …

Automatic/Scalable Firmware And System Software Update On Datacenter Products Like Server, Storage And Network Devices, Santhosh S, Deepak Acharya, Praveen Vishwakarma, Susanta Adhikary

Defensive Publications Series

The invention makes the Datacenter intelligent and scalable in case of “Firmware and System Software updates”.

Datacenter ecosystem generally consists of following major components

• Datacenter Products (Servers, Storage and Network devices)

• Central Management System (Manages the lifecycle of Datacenter Products).

• Administrators (Humans maintaining the datacenter)

  The innovation introduces a new way to do Firmware and System software upgrade and fresh installation on Datacenter Products. This is achieved with two newly introduced software:

• Smart Update Module

• The Store

  Smart Update Module:

  The Smart Update Module is intelligent software running on a Management Processor. Management processor is small integrated chip inside each …

  ---

  Go to article

Detection Of Message Read Status, Yan Liu

Defensive Publications Series

In some communication systems, notifications or other indications can be displayed to indicate whether received messages have been read by a user. In some cases, a read / unread status of a message, and display of this status, is inaccurate when that message has been received out of order, due to use of conversation timestamps by the system. This document describes techniques to accurately detect read and unread status for messages and to provide displayed indications of the status on all of a user's devices. This is achieved by storing an observed flag associated with each message that indicates the …

Intelligent Radio Search, Victor Carbune

Defensive Publications Series

Radio search using traditional techniques such as tuning or switching between pre-stored radio stations is cumbersome and time-consuming. It is difficult for users to locate a radio station that is currently playing programs that are of interest to them. The techniques described in this disclosure employ voice recognition, audio fingerprinting, and on-the-fly recognition to identify available content. Further, the available content is matched with known user preferences, based on user consent. An intelligent, assistive user interface based on matched radio content is provided to the user. The interface enables the user to easily find and listen to radio content of …

Mount, Tristan Gfrerer, Katie Roberts-Hoffman

Defensive Publications Series

This disclosure relates to a mount that allows a user to attach an electronic device such a camera, speaker, set top box, or specialized video conferencing system to another electric device or fixture. The mount may be affixed under a screen, to a wall, on the floor, or behind a screen inside a wall. The mount may include a screw hole for use with a screw to attach the mount to an electronic device. A user may insert a device such as a camera inside the U-shaped outwardly protruding portions of the mount prior to mounting. The user may remove …

Trust Chain Recovery Method For Secured Connected Devices Solutions, Olivier Kreet, Georges Nogueira

Defensive Publications Series

Computer networks are facing unprecedented security risks. Dramatically impacting threats are now targeting our critical infrastructures, such as :

• massive denial of services in the Internet because of lacking security in the Internet of Things (IoT),
• or, the recent ransomware campaign that spreads all over the world.

The need of security mechanisms is now inevitable to maintain the system. This target can be performed thanks to a set of now mature security protocols, but it will only succeed if we achieve the required security management of the multiplying seamless endpoints that do not allow any human interaction.

This solution addresses …

A System For Automatically Enhancing Advertiser Provided Sitelinks With Relevant Creative Text, John D. Lanza, John D. Lanza

Defensive Publications Series

Systems and methods described herein allow for providing enhanced sitelinks in the voice activated computer environment. A data processing system can determine a primary search result and a URL responsive to the identified user request. The data processing system can determine a primary search result and a URL responsive to the identified user request. The data processing system can identify a plurality of candidate digital components associated with the URL, and calculate a respective score for each of the identified candidate digital components. The data processing system can select a digital component of the plurality of candidate digital components based …

Dynamic Context-Based Voice Modulation, John D. Lanza, John D. Lanza

Defensive Publications Series

When audio sponsored content is presented with non-sponsored audio content in a two-way conversation between a user and a respective client computing device, it may be challenging for the user to distinguish the sponsored audio content from the non-sponsored audio content. The client computing device can receive an input audio signal indicative of a respective uttered speech. A data processing system communicatively coupled to the client computing device can process the received audio signal to identify a request corresponding to the uttered speech. The data processing system can generate a response to the identified request, and select an ad based …

Actions On Landing Pages, John D. Lanza, John D. Lanza

Defensive Publications Series

Systems and methods described herein allow for using audio overlay to provide actionable audio extensions with primary audio content. A data processing system can receive a first input audio signal from a client device of a user, and identify a user request from the first input audio signal. The data processing system can generate an audio response to the user’s request, and one or more audio extensions for presenting with the audio response. The data processing can then transmit the audio response and the audio extensions to the client device. The data processing system may receive a second audio signal …

Quality Assessment Signals, John D. Lanza, John D. Lanza

Defensive Publications Series

Systems and methods described herein allow for monitoring and adjustment of the quality of conversations between users and a virtual personal assistant. A data processing system hosting the virtual personal assistant can include a feedback monitoring component configured to determine parameters related to (or characteristics of) communication links between the data processing system and the client device, speech recognition and natural language processing of received audio signals, and/or user online activities responsive to content from the data processing system presented to the user. The data processing system can determine one or more quality metrics based on the collected parameters or …

Direct Actions, John D. Lanza, John D. Lanza

Defensive Publications Series

Systems and methods described herein allow for monetization of a virtual personal assistant by offering sponsored content to users responsive to input audio signals triggering online actions. A data processing system can identify a user request based on a received input audio signal. The data processing system can execute an online action responsive to the identified user request. The data processing system can generate a response to the input audio signal based on the executed online action, and select a sponsored content item based on a context of the identified user request. The data processing system can then transmit the …

Enabling Functionality In Content Items Based On Device Capabilities In Voice Activated Computing Systems, John D. Lanza

Defensive Publications Series

This document describes a technique for processing voice commands in a voice activated computing system. In particular, the system processes the voice commands and selects content to be sent to the user based on the capabilities of the user’s device. The system can poll the user’s device to determine their capabilities, such as for receiving voice commands, rendering audio responses, making phone calls, launching navigation application, accepting input via touch interface, etc. Based on the capability of a user’s device, the system may select content that can be best used to provide the user with the requested information or to …

Cost Per Second (Cps) Ads On Webpages, John D. Lanza, John D. Lanza

Defensive Publications Series

Systems and methods described herein allow for optimization of selection and monetization of ads for presenting during two-way conversations between users and a virtual personal assistant. A data processing system can receive an audio query from a client device and identify a user request based on the received audio query. The data processing system can determine a thread of tasks associated with the identified user request, and determine predicted segment durations for the identified tasks. The data processing system can identify candidate ads for presenting during a given segment, and determine a revenue value for each candidate ads. The data …

Bidding Per Viewable Impression, John D. Lanza

Defensive Publications Series

Systems and methods described herein allow for audio ad valuation based on respective historic presentation performance on client computing devices. The client computing devices can monitor presentation of audio ads to determine whether an audio ad was played and for how long. The client computing devices can report historic presentation performance values of ads to a data processing system. The data processing system can use the received audio ad presentation performance values to create and continually update a presentation probability predictive model. The presentation probability predictive model can be configured to predict probabilities of audio ads provided to client devices …

On-Demand Security Token Linking, John D. Lanza

Defensive Publications Series

Systems and methods described herein allow for automatic execution by a virtual personal assistant of secure access tasks, such as user account authentication or acquiring/granting access authorization, associated with an online service or online action requested by a user of a client computing device. The client computing device can receive an input speech signal indicative of a request or command from a user of a client computing device. A data processing system communicatively coupled to the client computing device can identify, based on the input speech signal, a service or online action requested by the user of the client computing …

Language Runtime Non-Volatile Ram Aware Swapping

Defensive Publications Series

One possible way to take advantage of NVRAM’s lower latency is to use it as swap space for a program. Moving stored data from RAM to NVRAM frees up RAM space that can then be used for storing other program data or instructions. Since data stored in RAM can be accessed faster, moving stored data from RAM to NVRAM increases the overall performance of the program. As an example, a language runtime may take advantage of NVRAM by identifying program memory that is less likely to be used in the near future and storing the identified program memory in NVRAM …