Punctuated Data Streams - Computer Action Team

Punctuated Data Streams - Computer Action Team

Data Models for Superimposed Information David Maier Maseeh Professor of Emerging Technologies Department of Computer Science Portland State University February 6, 2020 ER 2005: Data Models for SI 1 Collaborators

Concept & Architecture: Lois Delcambre, Sun Murthy Corporate Information Systems: Lougie Anderson, Radhika Reddy Clinical Information: Paul Gorman, Joan Ash, Mary Lavelle, Jason Lyman Forest Management: Tim Tolle Tobacco Documents: Jim Pankow, Patty Toccalino SLIM & SLIMPad: Shawn Bowers, Longxing Deng, Mat Weaver SIDEWALK (NSDL): Lillian Cassel, Ed Fox, Eric Hanson, David Archer February 6, 2020 ER 2005: Data Models for SI 2

My Austrian Connection Phillip von Jethon (Lemburg [Lviv], part of Austro-Hungarian empire in 1800s) Phillip Peter von Jethon (d. Stryj, 1881) Maria von Jethon Maier (Galicia, 1868) Jakob (James) Philip Maier (Bessarabia, 1899) Eugene Alfred Maier (Oregon, 1929) David Eugene Maier (Oregon, 1953) February 6, 2020 ER 2005: Data Models for SI

3 What is Superimposed Information? Information added over a base layer of documents and data to aid in re-finding, reuse, explanation, analysis, repurposing, reorganization. February 6, 2020 ER 2005: Data Models for SI 4 Superimposed Information has Existed for Millennia

Concordances Religious and legal commentaries Interlinear text Card catalogs Annotated bibliographies February 6, 2020 ER 2005: Data Models for SI 5 Concordance Index of occurrences of a term or phrase in a corpus, often with context Much Ado about Nothing [I, 3] 372 Don John Will it serve for any model to build mischief on? Pericles [II, 2] 759 Simonides It's fit it should be so; for princes are A model which heaven makes like to itself: As jewels lose their glory if neglected, So princes their renowns if not respected.

Hamlet [V, 2] 3701 Hamlet I had my father's signet in my purse, Which was the model of that Danish seal; February 6, 2020 DOMESTIC d., kind, and submissive to their Tit 2.05 DOMINEERING not as d. over those in your charge 1Pe 5.03 DOMINION let them have d. over the fish of Gen 1.26 and have d. over the fish of the 1.28 are you indeed to have d. over us? 37.08 By Jacob shall d. be exercised, and Num 24.19 the Philistines had d. over Israel. Ju 14.04 ER 2005: Data Models for SI 6

Commentary Explanation or analysis of a text, often piece by piece, sometimes with excerpts CHAPTER 2 Ge 2:1. THE NARRATIVE OF THE SIX DAYS' CREATION CONTINUED. The course of the narrative is improperly broken by the division of the chapter. 1. the heavens--the firmament or atmosphere. host--a multitude, a numerous array, usually connected in Scripture with heaven only, but here with the earth also, meaning all that they contain. were finished--brought to completion. No permanent change has ever since been made in the course of the world, no new species of animals been formed, no law of nature repealed or added to. They could have been finished in a moment as well as in six days, but the work of creation was gradual for the instruction of man, as well, perhaps, as of higher creatures (Job 38:7). Ge 2:2-7. THE FIRST SABBATH. 2. and he rested on the seventh day--not to repose from exhaustion with labor (see Isa 40:28), but ceased from working, an example equivalent to a command that we also should cease from labor of every kind. February 6, 2020 ER 2005: Data Models for SI

7 Legal Commentary Generally refers to laws or previous cases Although 66.0501(2), Stats., requires only that a governing body member resign before being appointed in order to be eligible to hold an office or position the selection to which is vested in the governing body and which was not created during the member's term, a governing body member should resign prior to applying to ensure that he or she does not violate 946.13(1)(a) and 19.59(1)(a) Stats. 10/29/01. Note reports U.S. Supreme Court decision to address how much protection the Fifth Amendment's public use requirement provides for individuals whose property is being condemned for the purpose of economic development in the eminent domain case from the Connecticut Supreme Court, Kelo v. City of New London, 268 Conn. 1, 843 A.2d 500 (Ct. 2004) and two eminent domain decisions from state courts also addressing the "public use" requirement, Bailey v. City of Mesa, 206 Ariz. 224, 76 P.3d 898 (App. 2003) and County of Wayne v. Hathcock, 2004 WL 1724875, ___ Mich. __, 684 NW.2d 765 (2004). 9/30/04. February 6, 2020 ER 2005: Data Models for SI

8 Interlinear Text Annotation interspersed with text. 12) Now when Jesus heard that John had been arrested, He left [the province of Judea] and went to Galilee. 13) [After] leaving Nazareth [i.e., a major town in Galilee], He came to live in Capernaum, [a city] beside Lake Galilee, near the districts of Zebulun and Naphtali. 14) [This was] so that what Isaiah the prophet had spoken would be fulfilled, saying, 15) [Isa. 9:1-2] The districts of Zebulun and Naphtali, near the lake, beyond [i.e., where it enters] the Jordan in Galilee, [the territory] of the Gentiles; 16) the people who lived [there] in [spiritual] darkness saw a great light. Light shone upon these people, who lived in an area that was under the shadow of death. February 6, 2020 ER 2005: Data Models for SI 9 Aside: Addressing Notice that some of these varieties of

information make use of a standard addressing scheme on the base layer. Book, Chapter, Verse Play, Act, Scene, Line Section, Subsection, Paragraph, Subparagraph February 6, 2020 ER 2005: Data Models for SI 10 Less Formal SI: Annotating* Whats going on here? Highlighting Commenting Classifying Grouping

Linking * Picture courtesy of Prof. James Pankow, Dept. of EBS, OGI February 6, 2020 ER 2005: Data Models for SI 11 A Bundle of Excerpts: Is This Certainly makes selections, adds structure SI? February 6, 2020 ER 2005: Data Models for SI 12 SI Has Carried over to Digital

World Bookmarks or Favorites in web browser A personal collection of locations in the World-Wide Web Simple interface: add & browse XLinks between web locations Comments feature in document applications Such as notes Add, move, delete, check-off Sort, summarize February 6, 2020 ER 2005: Data Models for SI

in Acrobat 13 Superimposed Applications Software tools that make use of SI Can be a separate interface or extend a base application February 6, 2020 ER 2005: Data Models for SI 14 NVivo Coding for qualitative analysis

Nodes associated with one or more text regions, can be organized in trees and cases Can add attributes to nodes and documents Annotate text with DataBites Can link out to segments of multimedia Create arbitrary sets of nodes and documents February 6, 2020 ER 2005: Data Models for SI 15 February 6, 2020 ER 2005: Data Models for SI 16

Some Superimposed Applications Weve Built Thingifier: Attach labels to things on the web February 6, 2020 ER 2005: Data Models for SI 17 February 6, 2020 ER 2005: Data Models for SI 18 Strand Map Browser Show dependencies among topics in a body of knowledge

Weve added connections to related materials Using to organize course units February 6, 2020 ER 2005: Data Models for SI 19 February 6, 2020 ER 2005: Data Models for SI 20 February 6, 2020 ER 2005: Data Models for SI

21 Superimposed Schematics Browser* Use ER model to identify and browse document collection for particular domain A kind of superware can be customized to different domains Example on US Forest Service appeals process *Developed originally by Shawn Bowers February 6, 2020 ER 2005: Data Models for SI 22 February 6, 2020 ER 2005: Data Models for SI

23 SLIMPad Trying to create a digital version of a clinical bundle February 6, 2020 ER 2005: Data Models for SI 24 Sidepad Demo Motivated by bundle building behavior Successor to SLIMPad and RIDPad Simple information model

Items with labels that mark into a base documents Groups of items and groups Labeled links (also with marks) Also support for accessing content and context of a marked region February 6, 2020 ER 2005: Data Models for SI 25 Data Modeling for SI

What are the main modeling concerns of SI? Can you separate superimposed information model from user interface from base layer access? What are the options for marks? How do you adapt your favorite data model for use as SI? How to cope with multiple models Querying across models Change of models February 6, 2020 ER 2005: Data Models for SI 26 Superimposed Information Model

Superimposed model can be different from base layer (which might be relatively unstructured) XLink: labeled graph (or hypergraph) Acrobat notes: location-comment pairs NVivo: nodes w/ multiple marks, w/ attributes, organized in trees and sets Thingifier: labeled collections (for classification) Strand Map Browser: Directed acyclic graphs (DAGs) Schematics Browser: Instance of ER schema with a mark on each entity or relationship instance (and a Mark value type) February 6, 2020 ER 2005: Data Models for SI

27 Sidepad Superimposed Model * Contains Group SidepadDoc Name * 0..1 ID Name Size Location 0..1

0..1 0..1 Belongs to * Item * February 6, 2020 ID Name Description Size Location Mark ID Address ER 2005: Data Models for SI

28 Representation How is information model stored Native representation Encoded into existing model: XML, Relational API could hide representation Decomposed into basic units: triples, RDF Flexible representation defined by metamodel Uni-Level Description (ULD, Bowers & Delcambre) February 6, 2020 ER 2005: Data Models for SI 29 Stand-off vs. In-line

Approaches Modify base layer versus reference it Issues with Stand-off Coping with changes in base layer Traversing from base to superimposed layer Issues with In-line May not own base layer Alternative SI on same base document How do you have SI span multiple documents?

February 6, 2020 ER 2005: Data Models for SI 30 Granularity of Addressing What are identifiable elements in base layer? Bookmark list: whole web pages (or anchor points provided by content) Sidepad: regions selectable by base application XLink: Whatever XPath can point to Interlinear text: Span of what an annotation refers to isnt always clear February 6, 2020 ER 2005: Data Models for SI 31

Embedded vs. Encapsulated Marks CARTE had mark-description info embedded directly in superimposed information (CARTE was our first superimposed application, adapted from Topic Maps) Hard to share marks across different SI models Sidepad has a repository of marks SI models use handles to them February 6, 2020 ER 2005: Data Models for SI

32 Semantics of Marks Who interprets marks Superimposed application Base application External agent (XPath in XLink) What can you ask about marks? Identity, Equivalence

Containment, Overlap Whats in it (content) Whats around it (context) February 6, 2020 ER 2005: Data Models for SI 33 Our Approach: SPARCE Superimposed Pluggable Architecture for Context and Excerpts Support multiple SI models Stand-off SI Address base layer at subdocument granularity Address space defined by base application Marks created in those applications Encapsulated marks, interpreted by base applications

Reactivate, Content, Context February 6, 2020 ER 2005: Data Models for SI 34 SPARCE Architecture Overview Superimposed Layer Middleware Base Layer Sidepad SPARCE Acrobat

XML Marks Schematics Browser Word Relations February 6, 2020 ER 2005: Data Models for SI 35 Excerpts and Contexts Name Value Excerpt

Finding of No Significant Impact not be necessary Font name Times New Roman HTML February 6, 2020

Finding of No Significant Impact

Based on ...

Excerpt

is the content of a marked region Context is information related to a mark A mediator called context agent retrieves content & context of a mark ER 2005: Data Models for SI 36 SPARCE Reference Model Superimposed Application Context Manager Base Application

Base Info Mark Manager Superimposed Info Mark Repository February 6, 2020 ER 2005: Data Models for SI 37 Different Approaches to Marks Are marks held in a common repository, or distributed among different superimposed applications?

Common Repository Provides sharing across different superimposed applications and superimposed document instances Provides collections for querying But binds superimposed document to a repository February 6, 2020 ER 2005: Data Models for SI 38 Sparcels Self-carrying code for marks sparce:markfodder=SUVBZ2VudHkJ Encodes application, document, region, user

Still need access to mark manager, context manager Need to be able to register a handler for the sparce namespace Potentially portable, but embeds addresses of documents February 6, 2020 ER 2005: Data Models for SI 39 API to Marks Can be through standard PL bindings, for example COM objects from Visual Basic or C++ Can provide services to activate base application, extract content, context, (update?) Also can be through a URI

Can provide different services via query part sparce://datalab.cs.pdx.edu/ markid=WordMark2003Apr24071826TYEEsmurthy?action=goto February 6, 2020 ER 2005: Data Models for SI 40 Where is Interpretation Done? Local server Requires software to be present locally Remote rendering service Needs to reformat base document to show highlighting Cant invoke local base application

Remote script service Returns a script to do local activation Tends to be browser specific February 6, 2020 ER 2005: Data Models for SI 41 Translating Marks Sharing establishing mapping between different user perspectives Your T: drive is my U: drive What about moving a mark between

Word, PDF and XML versions of a document? February 6, 2020 ER 2005: Data Models for SI 42 Extension vs. Intension Consider a patent as it undergoes revision: claims are modified, dropped and added The browser extension claim might be Claim 6 as submitted, Claim 5 in a revision and Claim 3 in the awarded patent. Want a way to talk about the same Claim across versions of the patent February 6, 2020 ER 2005: Data Models for SI

43 Semantic Marks A semantic mark is an excerpt (marked region) in a base document, an address for that marked region, and a user-defined label (possibly typed) that expresses the meaning of that marked region in a particular perspective, and

can be used to refer to the marked region Semantic marks allow referencing of marked regions based on meaning (content) rather location February 6, 2020 ER 2005: Data Models for SI 44 Semantic Marks A single label can reference multiple instances of the same concept When knowledge, belief, or the base document changes, can remap label to

a new marked region The history of mappings provides provenance a for concept February 6, 2020 ER 2005: Data Models for SI 45 Manual Semantic Marks Sticky tabs indicate the vertical position on the page Labels indicate the type of data being marked and provide an informal schema for organizing information February 6, 2020

ER 2005: Data Models for SI 46 Close-up One label summarizes multiple data items February 6, 2020 ER 2005: Data Models for SI 47 Another Close-up Emergent semantics? As-you-go schema? February 6, 2020

ER 2005: Data Models for SI 48 Superimposing Your Favorite Data or Document Model Want to take some existing model, and put marks in it If URL aware, can just use sparce namespace Word Excel HTML GetSmart (concept-mapping tool) Must rely on applications choice of where URLs may appear

February 6, 2020 ER 2005: Data Models for SI 49 Consider XML Could add extra sub-elements Extra sub-elements (hard for attributes) Alan Pratt How do you cloak added structure from tools that dont understand it? February 6, 2020

ER 2005: Data Models for SI 50 Could Use Parallel Structure Leave original as is Alan Pratt Build mark structure to be traversed in parallel February 6, 2020 ER 2005: Data Models for SI 51

What about Extending ER Model? Want to associate marks with any construct Entity Attribute Relationship Relationship attribute and use as an attribute value Can we do this without creating an entirely new model? E.g., come up with a convention to represent marking in an ER model February 6, 2020 ER 2005: Data Models for SI 52 The PC-Admin Browser

Allows a system (network) administrator to browse information about computers in a network Applications installed and the modules they use Updates applied Errors recorded Logged application, system, and security events User comments February 6, 2020

ER 2005: Data Models for SI 53 PC-Admin Mockup February 6, 2020 ER 2005: Data Models for SI 54 A Conceptual Schema* Observation Date Time Text User Relates to Computer Name

Error Date Time Source Description Notes Relates to Update Title Description Reason Applied on Applies to Occurs on Logged on Event

Date Time Kind Source Description Date Time Runs on Applies to I nvolves Application Name Uses Module Name

* All entities have key attribute ID (not shown); all relationships are many-many February 6, 2020 ER 2005: Data Models for SI 55 Base Layer: Event Log Date Time Source February 6, 2020 Description ER 2005: Data Models for SI Event Date Time Kind Source Description

56 Base Layer: Error Reports Date Time Description Error Date Time Source Description Notes Uniform structure, but mapping is not clean: Date and Time are both in Time field February 6, 2020 ER 2005: Data Models for SI 57 Base Layer: SW Updates

Title Description Update Title Description Reason Reason Data is heterogeneous and distributed: some data in XML, some in HTML Structure varies: Support URL not always defined, HTML page structure varies widely February 6, 2020 ER 2005: Data Models for SI 58 This is a Job for SI!

Heterogeneous base models Distributed sources Various base data are interconnected Logs are local, support pages are on the web Outlook errors stopped after SP2 was applied Conceptual schema hides heterogeneity and distribution, can represent connections Would like to navigate base layer through it February 6, 2020

ER 2005: Data Models for SI 59 How Should We Proceed? Want a methodology for extending ER model with references to marks Consider some alternatives February 6, 2020 ER 2005: Data Models for SI 60 Model Use of Mark as a Relationship? The traditional relationship does not completely capture the semantics of a mark Need to distinguish inter-layer and intra-layer relationships

February 6, 2020 ER 2005: Data Models for SI 61 ER Relationships Require Entities Might want to mark an attribute (e.g., Update.Title) Promoting attributes to entities, to attach relationships, reduces comprehension A group of attributes can share a mark (e.g., Error.Date and Error.Time)

Can be hard to find a key for an entity created for a group of attributes February 6, 2020 ER 2005: Data Models for SI 62 Marks in Attribute Values? Introducing a new domain Mark does not suffice to distinguish between mark and content-of-mark as values February 6, 2020 ER 2005: Data Models for SI 63 Supported Relationships

Some relationships have support An error applies to an application based on information in the details of the error report Error Date Time Source Description Notes A pplies to Application Name

Could use a relationship attribute, would need to know it marks relationship February 6, 2020 ER 2005: Data Models for SI 64 Our Approach Represent the use of a mark as a relationship Use relationship patterns to represent the use of marks February 6, 2020

ER 2005: Data Models for SI 65 Relationship Patterns A relationship pattern is an abstraction of recurring need when establishing relationships in a context Similar to a software pattern Specify signature, constraints Have standard translations to logical models February 6, 2020 ER 2005: Data Models for SI

66 Example: The Predicated Relationship Pattern Aircraft ID Range Route Route ID ID Assign (Range > 1.25*Distance)

Range Assign Distance ID Distance Aircraft signature: () is name of a relationship type; is a pre-condition for a relationship instance E.g., An aircraft can be assigned to a route only if it can fly at least 25% farther than the routes distance February 6, 2020 ER 2005: Data Models for SI 67 Modeling Marks

Mark ID The Mark entity models a mark The use of a mark is shown as a relationship with this entity All inter-layer relationships are between a superimposed entity and the Mark entity February 6, 2020 ER 2005: Data Models for SI 68

The Entity-Mark Pattern Event Computer Name Logged on Date Time Kind Source Description EMark:EventDetail 1 Mark ID EMark namespace for relationship types that associate entities with marks

EventDetail associates an Event entity with a mark February 6, 2020 ER 2005: Data Models for SI 69 The Attribute-Mark Pattern AMark:ErrorTime(Date, Time) Computer Name Error Date Occurs on Time Source Description Notes 1

1 AMark:ErrorDetails(Description) Mark ID AMark namespace for relationship types that associate attributes with marks ErrorDetails associates the Description attribute with a mark ErrorTime associates attributes Date and Time with one mark February 6, 2020 ER 2005: Data Models for SI 70

Deriving Attribute Values from Marks Update Title Description Reason VAMark:UpdateDetail(Description) 1 Mark ID An attribute might derive its value from a marks context VAMark namespace UpdateDetail associates the value of the attribute Description with the context of a mark February 6, 2020 ER 2005: Data Models for SI

71 The Relationship-Mark Pattern Computer Name Applied on Update Title Description Reason Applies to Application Name RMark:Application Mark ID

Aggregate* the relationship (called supported relationship) Add an RMark relationship for the aggregate First aggregate the AppliesTo relationship type. RMark:Application associates the aggregate with marks * Ramakrishnan and Gehrke. Database Management Systems, 3rd Ed. February 6, 2020 ER 2005: Data Models for SI 72 Visual Shorthand Computer Name

Applied on Update Title Description Reason Applies to Application Name RMark:Application Mark ID February 6, 2020 ER 2005: Data Models for SI 73

Revised Conceptual Schema RAMark:UpdateLog(Date, Time) Date Time Observation Date Time Text User Relates to Computer Name Applied on Mark ID

Update Title Description Reason VAMarks:UpdateDetail((Title), (Description), (Reason)) Applies to Occurs on Logged on Runs on Event Date EMark:EventDetail Mark Time ID Kind Source Description

Relates to Error AMarks:Error((Date, Date Time), (Description)) Time Source Description Notes Involves Applies to Mark ID RMark:Applicati on Application Name Uses

Module Name February 6, 2020 ER 2005: Data Models for SI 74 Dealing with Multiple Models Plethora of Models SI model and one or more base layer models Different SI models for same base layer Different models for different parts of SI? Want to move SI into another model (say for use with a different tool) Sidepad GetSmart

ULD helps with some of these Look at bi-level querying as a complementary technology February 6, 2020 ER 2005: Data Models for SI 75 Bi-Level Queries Ask queries over the superimposed information plus the content and context of marked regions Locate marked information based on its connections Select and format information for display Use superimposed information as lever to

do database style transformation and restructuring on non-DB information Build a common logical model for query Have talked about dual-language queries February 6, 2020 ER 2005: Data Models for SI 76 Basic Bi-level Query System Superimpose d Info Mark Info RIDPad Document,

XML PDF docs Word docs Property sets Transformers Context Agents XML XPath, XSLT, XQuery

Query Processor XML, HTML, February 6, 2020 ER 2005: Data Models for SI 77 Whats Neat about Bi-Level Queries Only mediate what we use (unlike

approaches that mediate entire sources) Works with different structures of superimposed information Useful in transforming and repurposing documents Adding a little bit of structure lets us do a lot of querying February 6, 2020 ER 2005: Data Models for SI 78 Example Query Retrieve all errors MS Word caused in the last week SELECT * FROM Error WHERE EDate BETWEEN CURRENT_DATE AND CURRENT_DATE - INTERVAL '6' DAY

AND Description LIKE '%Word.exe%' Error is a view: the attribute description is retrieved from the base layer when the view definition is executed February 6, 2020 ER 2005: Data Models for SI 79 Another Example Query Create a timeline of errors related to MS Word and MS Outlook SELECT EDate, ETime, Description FROM Error WHERE Description LIKE '%word.exe%' OR Description LIKE '%Outlook.exe%' February 6, 2020

ER 2005: Data Models for SI 80 Sample Results EDate ETime Description 1/26/2004 1/27/2004 3/9/2004 4/13/2004 4/23/2004 5/21/2004 5/26/2004 19:46 20:04 16:38

10:11 13:04 9:39 14:05 Hanging appOutlook.EXE Faulting appwinword.exe Hanging appwinword.EXE Faulting appOutlook.EXE Hanging appOutlook.EXE Faulting appwinword.exe Faulting appwinword.exe February 6, 2020 ER 2005: Data Models for SI 81 Result as Timeline* *Drawn using an XML transformation based on work of Nicolas Kruchten. Timeline is non-linear February 6, 2020

ER 2005: Data Models for SI 82 Summary Superimposed information has been around almost as long as writing Wealth of digital information presents even more opportunities for SI Added structure Information bridging Virtual documents

Interesting modeling issues Whats a mark? Adapting a data model for SI Moving between SI models February 6, 2020 ER 2005: Data Models for SI 83 Questions? http://datalab.cs.pdx.edu/sparce/ February 6, 2020 ER 2005: Data Models for SI

84 My Question Which is furthest east? Berlin Naples Valletta Klagenfurt February 6, 2020 ER 2005: Data Models for SI 85 Back-up Slides

February 6, 2020 ER 2005: Data Models for SI 86 Demo Group Item February 6, 2020 ER 2005: Data Models for SI 87 Creating a Mark February 6, 2020 ER 2005: Data Models for SI

88 Creating a Mark February 6, 2020 ER 2005: Data Models for SI 89 Creating a Mark February 6, 2020 ER 2005: Data Models for SI 90 Transformation via Query February 6, 2020

ER 2005: Data Models for SI 91

Recently Viewed Presentations

  • Diapositive 1 - Free

    Diapositive 1 - Free

    Polytech'Orléans Filière ESI MODULE FILTRAGE COMPRESSION FASCICULE DE COURS Filtrage Multicadence ANNÉE 2006-2007 SPE 4 Dr. Rodolphe WEBER Multicadence - R. WEBER - POLYTECH'ORLEANS -06/07 Conception de filtres numériques Changement de Fréquence Filtrage multicadence Analyse/synthèse de signaux Compression de signaux...
  • Old Testament Survey: Books of Leviticus and Numbers (22)

    Old Testament Survey: Books of Leviticus and Numbers (22)

    Old-Testament Survey: Books of Leviticus and Numbers (22) Laws Governing the Priesthood and Census * * * * * * Leviticus Third book written by Moses; means "pertaining to the Levites." Records the work of the Levites appointed to serve...
  • Distance Education in the SDICCCA Region: 2016/17 DE ...

    Distance Education in the SDICCCA Region: 2016/17 DE ...

    Canvas mentor training fall 2016 within disciplines to assist with the transition. Canvas Pilot Started Spring 2017 with about 8 faculty . Continuing to offer Canvas training, 4 week online course, self paced course and Professional development workshops. Developed Student...
  • Machine Learning Lecture 4

    Machine Learning Lecture 4

    A few things.. If you haven't starting using GPU server, please do it now! Kaggle competition ends in less than 1 month, please start it right now
  • Semester Project Underwater Optical Communication

    Semester Project Underwater Optical Communication

    Summary Goals Communication technology Optical Experiments Fast Optical Communication Transmitter Receiver Future work Goals Project Description Radio Blue light [receiver] [transmitter] video and data accompanying vehicle control Goals Goals of the Project Develop a communication system to transmit video between...
  • Foreign Aid, Foreign Policy and Development Management

    Foreign Aid, Foreign Policy and Development Management

    foreign aid, foreign policy and development management louis a. picard pia 2096/pia 2490 week six
  • COMA COMA CENTAURUS  CENTAURUS Hebrews 1:3 And He

    COMA COMA CENTAURUS CENTAURUS Hebrews 1:3 And He

    Gr. Noun, HUPOSTASIS ( (upostasis) the essential or basic structure/nature of an entity, substantial nature, essence, actual being, reality. John 1:1-2. In . the beginning was the Word, and the Word was with God, and the Word was God. 2)...
  • Storing Data: Disks and Files - site.uottawa.ca

    Storing Data: Disks and Files - site.uottawa.ca

    L'information redondante permet de reconstruire les données en cas de faillite d'un disque: Utiliser tous les disques (« striping ») vs. N'utiliser que des ''check disks'' pour l'info redondante. L'info redondante est calculée en utilisant un schème de parité.