Digital Commons Network^™

Automatically Scaling Multi-Tenant Machine Learning, Steven J. Ross, Christopher Olston, Noah Fiedel

Defensive Publications Series

Generally, the present disclosure is directed to optimizing use of computing resources in a system. In particular, in some implementations, the systems and methods of the present disclosure can include or otherwise leverage one or more machine-learned models to predict task allocation for a job serving a plurality of machine-learned models based on current system state and queries per second (QPS) data for the plurality of models. Alternatively, the tasks can be allocated according to one or more rules (e.g., a new task is allocated to a job until the compute usage for the job falls below a scaling threshold). …

Legal Precedent Mining With Machine Learning, Alexandros Salazar

Defensive Publications Series

Generally, the present disclosure is directed to gathering legal precedent from a corpus of law. In particular, in some implementations, the systems and methods of the present disclosure can include or otherwise leverage one or more machine-learned models to predict decision data from a court decision based on the court decision (e.g., a textual copy of the court decision).

Determining Validity Of A Point Of Interest Based On Existing Data, Melanie L. Donny-Clark, Jingyu Zhou, Shuang Qu

Defensive Publications Series

Generally, the present disclosure is directed to determining whether or not a point of interest is valid based on existing data. In particular, in some implementations, the systems and methods of the present disclosure can include or otherwise leverage one or more machine-learned models to predict a likelihood of a point of interest being valid based on ‘clues’ gathered from existing data.

Extracting Point Of Interest Information From Query Logs, Melanie L. Donny-Clark, Jing Dai

Defensive Publications Series

Generally, the present disclosure is directed to extracting information about a point of interest from one or more query logs. In particular, in some implementations, the systems and methods of the present disclosure can include or otherwise leverage one or more machine-learned models to predict local intent for a query based on content of the query.

Correcting Image Anomalies Using Machine Learning, Nick Felker

Defensive Publications Series

Generally, the present disclosure is directed to correcting anomalies in an image. In particular, in some implementations, the systems and methods of the present disclosure can include or otherwise leverage one or more machine-learned models to correct one or more anomalies in an image based on image data from an image.

Fitness Program For User-Defined Goals, Nick Felker

Defensive Publications Series

Generally, the present disclosure is directed to a fitness program tailored to a user. In particular, in some implementations, the systems and methods of the present disclosure can include or otherwise leverage one or more machine-learned models to generate a goal-oriented fitness program to a user based on user physiological and or fitness data.

User Interface And Application Programming Interface For Explaining Personalized Machine-Learned Model Outputs, N/A

Defensive Publications Series

Generally, the present disclosure is directed to systems and methods that enable explanation of personalized machine-learned model outputs. For example, the systems and methods of the present disclosure can include or leverage a user interface and/or an application programming interface (API) to enable explanation of model outputs. According to one aspect, the present disclosure provides a principled way of discovering which actions or range of actions that the user took contributed to a particular output that the model predicted at later time. This allows, for example, an application that employs the personalized model to describe to the user which actions …

Evaluate Presentation, Wendy Kozlowski, Heidi Imker

IASSIST & DCN - Data Curation Workshop

The Evaluate step to curate research data is explained in this presentation.

Optimizing Procedural Tov-112 Ovarian Carcinoma Cell Transfection Via Initial Seeding And Pbabe/Pei Jet Ratio Variations, Cassondra Walker

Honors Theses

Cell transfection is an exemplary tool in protein analysis and examining transcription factor activity. While general procedures are commonly followed, each cell line requires slight differentiation in pBABE/PEI JET ratios and initial cell count for optimal transfection results. This research focusses on determining a successful procedure for TOV-112 carcinoma cells for future work with 2-methoxyestradiol and NFkB. 2-methoxyestradiol is an angiogenesis inhibitor believed to decrease inflammatory processes. NFkB promoter controls the luciferase gene on the MetLuc plasmid and is pro-inflammatory. The objective of this experiment is to modify the variables until successful results at 24 and 48 hours are observed. …

Providing Reader Assistance Based On Reader Interaction, Nick Felker

Defensive Publications Series

Generally, the present disclosure is directed to assisting users of an electronic document. In particular, in some implementations, the systems and methods of the present disclosure can include or otherwise leverage one or more machine-learned models to provide assistance for one or more content items in an electronic document to a user of the electronic document based on the user’s interaction with the electronic document.

Machine Learning To Predict Advertisement Targeting Solutions, Zachary Garrett

Defensive Publications Series

Generally, the present disclosure is directed to using machine learning to predict advertisement targeting solutions. In particular, in some implementations, the systems and methods of the present disclosure can include or otherwise leverage one or more machine-learned models to predict optimal advertisement target solutions such as, for example, keyword word sets, negative word sets, location restrictions, bid adjustments, and/or schedules based on product data such as, for example, advertisement content (e.g., ad creatives text), seed keywords, images of the product, and/or advertiser metadata.

Using Image Recognition To Control Devices, Johan Ulrich Lewin Jessen, Morten Just Petersen

Defensive Publications Series

Generally, the present disclosure is directed to using image recognition to control devices. In particular, in some implementations, the systems and methods of the present disclosure can include or otherwise leverage one or more machine-learned models to predict an appropriate user interface to control one or more remotely controllable appliances based on image data.

Generating Icons For Applications In An Applications Marketplace, Jun Hong

Defensive Publications Series

Generally, the present disclosure is directed to generating one or more icons for one or more applications in an applications marketplace. In particular, in some implementations, the systems and methods of the present disclosure can include or otherwise leverage one or more machine-learned models to generate an icon for an application based on one or more icons for existing applications.

Transform Presentation, Lisa Johnston, Jennifer Moore

IASSIST & DCN - Data Curation Workshop

The Transform step to curate research data is explained in this presentation.

Classifying Aspect Ratios Of Images, Andrew Gallagher

Defensive Publications Series

Disclosed herein is a mechanism for detecting and/or adjusting aspect ratio in images and videos. For example, the mechanism can include analyzing a digital image having pixels by identifying sets of pixels corresponding to one or more objects in the digital image, analyzing the set of pixels corresponding to at least one object in the image to determine an estimated image aspect ratio, and applying an image transform to the digital image based on the estimated image aspect ratio to generate a modified digital image. In a more particular example, the one or more objects can correspond to faces, animals, …

Machine Learning To Disable Applications From Using Background Resources Except At Appropriate Times, Thomas Price, Eytan Lerba

Defensive Publications Series

Generally, the present disclosure is directed to disabling computing applications from using background resources (e.g., processing or data transmission) except at appropriate times (e.g., times when the user is likely to use the electronic device). In particular, in some implementations, the systems and methods of the present disclosure can include or otherwise leverage one or more machine-learned models to predict an appropriate time for an application to use resources based on device usage data.

Machine Learning To Select Best Network Access Point, Thomas Price, Eytan Lerba

Defensive Publications Series

Generally, the present disclosure is directed to selecting a best network access point when multiple are available (e.g., including cell towers, WiFi access points, etc.). In particular, in some implementations, the systems and methods of the present disclosure can include or otherwise leverage one or more machine-learned models to predict a best network access point based on device usage data.

Rethinking The Potential Of Documentation Of Culture As A Data Gathering Practice, Tomasz Umerle

Proceedings from the Document Academy

In this article, I examine the documentation of culture (DoC). This is the practice of gathering and producing the information and data relevant to intangible cultural phenomena. DoC is generally practiced by teams of documentalists generally outside GLAM institutions (i.e., galleries, libraries, archives and museums). As such, it differs from the preservation and description of concrete objects of cultural heritage done within those institutions. I examine how DoC finds it increasingly difficult to clearly define and communicate its role: a) in relation to LIS; b) to the broader academic community; and c) in digital age of information overload. In this …

The Value And Problems Of Digital Preservation For Historical Documents In China, Yongsheng Chen, Huanning Su

Proceedings from the Document Academy

China has been taking on large-scale digitization of historical documents. This paper summarizes four advantages of the digital preservation of historical documents from the practice of Chinese archives. First, it is good for protecting the originals; second, it makes the historical documents more convenient to use; third, it lays a resource foundation for the construction of digital archives; last, it is beneficial for utilizing historical documents. This paper also points out problems in China's approach to the digital preservation of historical documents. First, it has been proceeding without a plan. Such a plan should be created by following the Value …

Deconstructing The Visa: Academic Accommodations And Self-Disclosure, Catherine A. Smith Phd

Proceedings from the Document Academy

College students living with disabilities must disclose, manage, share, and communicate their personal health information in order to receive academic accommodations. The academic gatekeepers to this process are disability services specialists who are important “third parties:” non-family members, non-friends, who work with clinical information to meet the educational needs of their student clients. The author reports the results of the first of several studies of different stakeholders in the academic accommodations process and their attitudes about Personal health records (PHRs). Semistructured interviews were conducted between January and April, 2013, with 17 Disabilities Services (DS) specialists at 2 different public 4-year …

On The Borders Of The Document: Trip To Turakia, Sabine Roux, Caroline Courbieres

Proceedings from the Document Academy

Analysis of documentary boundaries through the study of a set of documents related to the French Theater Company Turak. We describe as artistic documents the different documents that circulate during the creation of a play. There is an initial documentary collection comprised of the documentation for the performance on which the company relies to stage the play itself. At different stages of artistic creation, the company designs documents that are useful for the pursuit of creative reflection. The theater company also develops documents aimed at promoting the play for professionals and the public. Finally, the spectators also produce many documents …

Representativity And Complementarity In Tai Chi As Embodied Documentation, Joacim Hansson

Proceedings from the Document Academy

The aim of this paper is to investigate what happens if we leave the criteria of materiality and permanence behind in the study of documents. How far can we stretch the definition of a document or define a documentation process in a situation where neither the originary fact, or object, nor that by which this is represented is material or permanent? Empirically, the paper is constructed as a case study of the traditional Chinese practice of Tai Chi and presents a formulation of the Tai Chi form as an immaterial document and Tai Chi pratice as a doumentation process. The …

[For The System, Alternate Title: If It Sort Of Looks Like A Duck: Reflecting On Bad Photographs And Chains Of Custody], Jodi Kearns, Brian C. O'Connor

Proceedings from the Document Academy

Though the system will not permit it, our abstract is an out-of-focus photograph of ducks at 1900 pixels wide and black and white, which is approximately 20% the size of the original color photograph we use for our title. By most technical standards, it is a bad picture. Straightening the horizon, cropping the image to emphasize the two foremost ducks, brightening the image to highlight the feet, and adding a caption that indicates activity might yield a “better” picture for some viewers. This piece captures nearly 20 years of conversations about good and bad pictures, and continues the conversation from …

Digital Film Preservation: The Search For The Original, Roswitha Skare

Proceedings from the Document Academy

No abstract provided.

Training The Masses In “Informational Awareness”, Carol Choksy

Proceedings from the Document Academy

The frameworks for “information literacy” and “media literacy” require a complex understanding of authority and context and neglect the most basic of ideas, that all sources have a perspective. Whether that perspective is useful or not is dependent upon its purpose. If a student is researching conspiracy theories, they will be looking at a vast array of perspectives. If a student is researching active measures, they will be looking at an even broader array. Beyond learning to determine whether a news article is “Fake News” students must be able to filter the snippets of information that inundate them on a …

Before The Antelope: Robert Pagès On Documents, Michael K. Buckland

Proceedings from the Document Academy

In 1951 Suzanne Briet wrote, with minimal explanation, that an antelope could become a document. In 1948 Robert Pagès (1919-2007) published an explanation of the same and related ideas. Textual and other graphic documents are about something, hence descriptive and derived. Animals and other objects are informative because they are illustrative of themselves either as specimens of a class (tokens of a type) or simply as particular individuals (“autodocuments”). Pagès’ career and ideas are briefly discussed.

Promoting Trustworthiness And Discoverability In An S-Town Of Osn Documents, Laurie J. Bonnici

Proceedings from the Document Academy

The realm of online social networks (OSN) has evolved rapidly over the last decade. While literature has primarily focused on Twitter, Facebook presents a unique forum for seeking secondhand knowledge on highly specialized topics including life-threatening medical conditions. Cognitive authority expressed through accounts of personal experiences augments authoritative resources that often are inaccessible or non-existent. OSN posts in Facebook groups appear as a plethora of documents thrown upon the virtual floor, disorganized and unsearchable. This paper offers ideas of what could be possible when interaction design (Ixd) is applied to Facebook groups to promote trustworthiness and discoverability for the information …

Science Serving Industry: Documentary Authority And Industrial Influence In 19th Century American Chemistry, Shawn Martin

Proceedings from the Document Academy

No abstract provided.

A Palimpsest Of Diné Voices, Frances Vitali, Brian C. O'Connor

Proceedings from the Document Academy

We use the idea of a palimpsest to probe and illuminate Hayse's model of communication as a dance - not simply the erasure of a single manuscript but the scraping away of the communication system of an entire people.

Among Navajo (Dine’), oral tradition and oral language still operates within mediated print and digital technologies as a complementarity. There is still a commitment to honoring whose voice is telling the stories in American Indian literature, for colonialist attitudes may still be coded with unchallenged stereotypes, cultural inaccuracies for readers.

Modeling the intersection of Diné culture and Anglo culture – both …

The Materiality Of Documents And The Genocidal Continuum, Martin I. Nord, D. Grant Campbell

Proceedings from the Document Academy

This paper uses Nancy Scheper-Hughes’s concept of the “genocidal continuum” to examine the materiality of documents in the context of Canada’s historical treatment of indigenous peoples. We discuss Canada’s Report of the Truth and Reconciliation Commission, using a hermeneutic perspective inspired by Hans-Georg Gadamer, and argue that the Report’s status as a material document plays a complex role in Canada’s acknowledgement of its past treatment of indigenous cultures and its current efforts at reconciliation. Virtually all societies contain philosophical, religious, and social norms that implore us to be responsible, ethical, or loving to everyone, even the stranger. But in cases …