Satisfied Sunday obligations at Saturday 5pm Mass, followed by dining out. Routine the girls & I have established. Actually, makes gr8 sense
Caught "Inside Llewyn Davis" on DVD. Liked it. Spare, taut, linear plot development. Coen Bros do good unfussy character study. Great music
Wardrobe malfunction. Malwear.
Good week. Fairly mellow. I don't use the word "mellow" much. But the week gone by sailed by on pure intution. For me, that's mellow.
Damien Jurado "Life Away From the Garden" (http://ow.ly/A8iLE ) JK--2012. From "Maraqopa," one of my fave LPs of the 2010s. Dig th chorus
When I write, little birdie in my head always asking "Jim, didn't you already say that somewhere?" Yes, birdie & I are on a first-name basis
"Open Data on Net Neutrality: Help Crowdsource Analysis o Comments to the FCC" (http://ow.ly/A6NOY ) JK--NLP data scientists: have at it!
"Data Science Cheat Sheet" (http://ow.ly/A6MZZ ) JK--Huh!? This is busy as all get-out. Cheatsheets are supposed to be simple cribnotes
Mary Tyler Moore Show "Ted's Moment of Glory" (http://ow.ly/A6LaV ) JK--Cut to 20:00 mark for classic of Lou taunting Ted: "QUIZmaster!"
"Securing Billions of IoT Devices Poses Mind-Boggling Challenges" (http://ow.ly/A6JV0 ) JK--Mnd-bgglng device #, diversity, connectedness
"Audience Dead, Long Live Active Content Consumrs" (http://ow.ly/A6Jjp ) JK--"Universal authorship": every1 entitled to 2-cents everythng
"Internet of Things: A Big #BigData Use Case" (http://ow.ly/A6Hi2 ) JK--EMA's JMyers says ROI calc 4 machine-data analytx straightforward
"How #BigData Helps Tiniest Patients" (http://ow.ly/A6GeM ) JK--Kaiser improved predictive risk assessmt o early-onset sepsis in newborns
"Spotify uses deep learn 2 improv playlists" (http://ow.ly/A6Dwn ) JK--Recomm algos drivn by song acoustic analytx find cool unheard jams
"Data Scientist Role Shifting to Focus on Developers" (http://ow.ly/A6CTV ) JK--Actually, data scientists ARE the new developers.
"Making #BigData Work: Retail Energy" (http://ow.ly/A6Cuj ) JK--Interesting graph forecasting amount o data to be generated by smarthomes
How Will the Internet of Things Evolve in the Next 100 Years? http://bit.ly/1r0a8cn pic.twitter.com/22KxA53Gdk
This wk #IBMDataMag qhs: http://ow.ly/A6zlN http://ow.ly/A6zmQ http://ow.ly/A6zoJ http://ow.ly/A6zrE http://ow.ly/A6zsP
Conversation optimization? ...accessorizing your lifestyle online p1: http://ow.ly/A6ydW p2: http://ow.ly/A6yfC Fri #IBMDataMag qh
"CSC #IBM Expand
Alliance to Speed Client’
Adopt o Hybrid Cloud Mobile #BigData" (http://ow.ly/A6tbI )
JK--CoE offers #BlueMix
RT @kexpplaylist #kexp Staircase at the University by Morrissey frm World Peace... JK--"If it breaks your legs then dont come running to me"
"New #IBM SyNAPSE Chip Could Open Era of Vast Neural Nets" (http://ow.ly/A6rrd ) JK--Cognitive brain-inspired non-von Neumann architcture
"10 most cringe-worthy tech movies ever made" (http://ow.ly/A6nYi ) JK--They forgot "War Games" (1983) with Matthew Broderick. Stinker
What's the deal with the hyphen in "Spider-Man." Is the hyphen intended to symbolize the spider web threads the man shoots from his wrists?
RT @kexpplaylist #kexp Thanks a
Lot by Neko Case & Her Boyfriends from The Virginian
JK--1997. She should do another traditional country LP
RT @kexpplaylist #kexp Harper Valley P.T.A. by Jeannie C Riley JK--1968. Great Tom T Hall composition that was a big crossover pop radio hit
Did my yoga alone out along the mirror wall in the big common room at LA Fitness. Kept my eyes closed. Self-consciousness screws me up.
New #IBM jk blog: "Wearables driving real-time actionable analytics into the modern lifestyle" (http://ow.ly/A5ALy )
GESU//They walk on water/like Jesus of Nazareth/if he were a goose.
OK, enough bigness for one day. Enough dataness too, while we're on the topic. Enough, already.
In response to yr many requests, no, I'm not on a first-name basis w/ #Watson. I doubt he even has a first name. I even doubt that he's a he
RT @data_henrik: A8: learn more about #fraud by posting your #SSN and credit card number here.... #FraudPrevention JK--LOL!!!!!!
A8: #FraudPrevention Here's a great #IBM resource to learn more about #banking anti-fraud issues & solutions: http://ibm.co/V4ONQ5
A7: #FraudPrevention New #bigdata techs--eg #Hadoop #NoSQL--support #banking fraud detection by execute analytx in scalabl MPP cloud pltfrms
A7: #FraudPrevention New techs such as in-memory, a la #IBM #BLU, help in #banking fraud detection through speed & tight OLTP integration
RT @wilsondavisibm: #FraudPrevention A6: Entity analytics very necessary part. List processing, relationship resolution, household analysis
#FraudPrevention A6 tight latency requires fast platforms, clean models, simplified business processes around all that
A6: #FraudPrevention Interdicting fraud in real-time requires lo-latency decision-automation closed response-loop back to banking OLTP apps
A6: #FraudPrevention Analyzing #banking transactions as occur requires serious lo-latency #bigdata analytic horsepwr, a la the #IBM platform
A6: #FraudPrevention Analyzing transactions as occur requires OLTP DBMSs with streaming & in-memory capabs: eg #IBM DB2 w/ BLU & Streams
A5: #FraudPrevention When talk turns to anti-fraud customer "profiling," privacy & discrimination sensitivities loom large. Tread carefully
A5: #FraudPrevention Banks should leverage third-party commercial data sources on customers to round out profiling of suspect segments
A5: Anti-Money Laundering Data Marts, Phone Support notes, Teller Information. #FraudPrevention
A5: #FraudPrevention Key #banking anti-fraud data sources: transactional records, security audit logs, & customer/account profiles.
A4: #FraudPrevention Transactional data mining & segmentation analysis are fundamental to distinguish fraudulent vs legit #banking activity
Nerds who tape on their broken glasses to keep them on their face? Anachronism! Modern nerds tape lazik-adjusted eyeballs to their sockets
A4: fraud model gets applied in real-time to current activities, matches are further analyzed / verified #FraudPrevention
A4: #FraudPrevention Require predictive propensity models & deterministic business rules to pre-empt & contain fraud before it can do damage
A4: #FraudPrevention Detect fraud requires historical trans analysis, pred models, stream analysis, social graph analysis, outlier detection
A3: #FraudPrevention Multifactor strong authentication flags can flag/deter fraud attempts. Analyze data to pinpoint+predict future attempts
A3: Predictive Fraud #Analytics, Social Network Analysis, Multi-factor Authentication & Cognitive Apps will all help. #FraudPrevention
A3: data protection, security, encryption help significantly #FraudPrevention
A3: #FraudPrevention Identifying anomalous banking transction patterns involvs per-customer historicl analysis & profiling 2 flag exceptions
A3: #FraudPrevention Identify potential fraudulent transctions by looking 4 anomalies: unusual amounts, times, locations, access chanls, etc
A2: #FraudPrevention Another key impact of #banking fraud is shift of biz resource allocation toward risk mitigation, away from productivity
#FraudPrevention- A2- Fraud false positives are a key challenge and impact customer satisfaction. #bigdata #fintech
A2: #FraudPrevention Another key biz impact of #banking fraud is widespread lack of trust in the online financial system.
#FraudPrevention-A2- A Harris Interactive poll in 2013 found 71% would switch banks if they were a victim of fraud. #fintech #bigdata
A2: #FraudPrevention Key biz impacts of #banking fraud are costs, lags, headaches, & complexities associated w/ensuring iron-clad security
A1: #FraudPrevention Under cover of strong commercial encryption, fraudsters can rob banks blind & remain hidden.
A1: #FraudPrevention Prevalence o zombies, bots, & worms in cyber-crime is shocking. Financial institutions everywhere under constant attack
RT @data_henrik: RT @wilsondavisibm #FraudPrevention A1: It's increasing. Now about 80% of #bank #fraud is from organized crime.
#FraudPrevention A1 in about 30% of cases, internal collusion is involved. (providing lists, data, information)
A1: #FraudPrevention In addition to new online channels for bank fraud, there are new online social/collaboration tools for fraudsters
RT @InfoMgmtExec: A1: Fraud always increases during Economic Downturns, but has really risen during this past 1. Epidemic. #FraudPrevention
RT @wilsondavisibm: #FraudPrevention A1: They are certainly increasing. From usually 1-2% of revenue to closer to 5%
Fraud always increases during Economic Downturns, but has really risen during this past one. It's an epidemic now. #FraudPrevention
A1: #FraudPrevention Criminals certainly have more channels for perpetrating bankd fraud these days: web & mobile. It can't be decreasing.
@IBM_@bigdatabusiness @wilsondavisibm DB2 Pleased to be participating in this #FraudPrevention tweetchat.
I will be taking part in #FraudPrevention in #Banking tweetchat today from 11-12 ET. Please join me: http://bit.ly/1kL1XOt #fintech
Advanced analytics? Pick an algorithm, any algorithm p1: http://ow.ly/A3Qm7 p2: http://ow.ly/A3Qne Thurs #IBMDataMag quick-hit
Just because I'm "following" somebody on socials, doesnt necessarily mean I'm reading them. But it does mean that I'm not "not reading" them
“Most biz only analyze 12% of their data.” InfoSphere BigInsights 4 Hadoop can change that! http://ibm.co/XAP0fz #bigdata
Cinematic title sequences are an art in their own right. Deserve own Oscar. Saul Bass did great ones in the 60s, eg: http://ow.ly/A3DUh
Does EU's so-called "right 2 B forgotten" expire at the subject's death? Can posterity petition to have historical record set right again?
Data "lake" doesn't make sense. Better 2 say data "reservoir": holds resource 2 B consumed l8r. Or data "sandbox": playpen 4 data scientists
Would it be tempting fate to include kicking an actual bucket on your bucket list?
Stereotypical geeks in clashing colors, dress-shirt pocket protectors, & Scotch-tape-repaird eyeglass frames. I've never seen one in th wild
I always GGL foreign words that natively hav accent marks not in English alphabet. I use GGL response text in my work. Eg, crème de la crème
"Gravy train"? At what point did "gravy" become synonym for "luxury"? Was it ever transported in bulk on the railroads like crude oil?
Are frauds in #banking increasing or decreasing? Join us Thu 8/7 at 11am EDT http://bit.ly/fraudChat #FraudPrevention #DB2
" #IBM Extends Consulting to New & Existing Clients with Click-to-Buy Services" (http://ow.ly/A1Obn ) JK--Analytic expertise as a service
Ask these 4 questions before attending any conference. For #ibminsight the answer is yes, yes, yes & yes. https://ibm.biz/BdFHSx #analytics
Participating in the IT industry standard-setting process is a bit like watching paint dry, but not as fun.
Join me, @bigdatabusiness, & @wilsondavisibm tomorrow #IBM tweetchat #FraudPrevention in Banking, 11am-12noon EDT (http://ow.ly/A1vMW )
VoltDB and #IBM Present: The Key to Unlocking the Economic Value of Fast Data in the Cloud http://bit.ly/1kHpik7 pic.twitter.com/718IM2ICf9
RT @graemeknows: "Talking to @jameskobielus. Feelin' the flow. Workin' it." JK--Graeme's one of the few I know who brainstorms fast & well
I always skip to byline of any tech article. Curious who author is, what their agenda might be. Then I skim 2 C how egregiously they push it
"Predictive Modeling w/ #BigData: Is Bigger Really Better?" (http://ow.ly/A1cDX ) JK--Research showing yield of fresh insights as U scale
"IBM Named Only Leader in Gartner MQ 4 Mobility Svcs" (http://ow.ly/A18TM ) JK--And GG did research BEFORE we announced Apple partnership
Catch in-person panel with me @sardire @ajbowles @n2semantics Thurs 8-21 San Jose Cognitive Computing Forum (http://ow.ly/A17Fj )
Decis automation? Retraining & restraining th long-data arm o the law p1 http://ow.ly/A12Jv p2 http://ow.ly/A12L5 Wed #IBMDataMag qh
"Direct Comparison of #IBM InfoSphere Streams & Apache Storm in A Real World Use Case" (http://ow.ly/A0Ydb ) JK--Good whitepaper
RT @kexpplaylist #kexp Got Nuffin' by Spoon JK--Live right now from Dexter&Denny. Standout cut from
Austin band's 2010 LP
"When data becomes dangerous" (http://ow.ly/zYL0K ) JK--Sci-fi-fueled freakouts not conducive to sober analysis
"Tyranny Of Th Beep: Taming Th @#$! Sensors" (http://ow.ly/zYcPa ) JK--Computer noise pollution kills spirit thru stream of auditory cuts
"Lab that created Spark wants to speed up everything" (http://ow.ly/zYbxc ) JK--Good one on UC-Berkeley AMPLab & Brkly Data Analytix Stak
"SQL-on-Hadoop: Full Circle Back to Shared Nothing Database Architectures" (http://ow.ly/zYaGm ) JK--Great #IBM research article.
"Proceedings of the VLDB Endowment" (http://ow.ly/zYacb ) JK--Unbelievably rich trove of #bigdata #datascience research journal articles
"Containers: The Enabler of YARN" (http://ow.ly/zY8Ji ) JK--Discusses Docker in #YARN #Hadoop context.
Reading Don Rickles' memoir. I'm struck by his utter boldness & counterintuitive approach to comedy. No jokes, just insults, & folks luv it!
"Computer Programming Is a Trade; Let's Act Like It" (http://ow.ly/zY421 ) JK--Exactly right. Univ degree not necessary 2 B gr8 developer
Join me, @bigdatabusiness & @wilsondavisibm for #IBM tweetchat #FraudPrevention in Banking, Thurs Aug 7, 11am EDT (http://ow.ly/zY1tD ).
I always make point of specifying precise state/city where DC-area event takes place. Especially if it's
We're NOT WashDC
"The Big Data Seminar 2014" (http://ow.ly/zXZZs ) JK--September 17–18,
Arlington VA, . Krishnan &
Inmon #IBMDataMag Hilton Crystal
RT @kexpplaylist #kexp
Thunder Road by Bruce Springsteen from
Born to Run JK--1975. Nicely shows off the rocking chops of entire E St Band.
"Take a Quantitative Look at Insight 2014" (http://ow.ly/zXYjQ ) JK--October 26-30, Vegas. Register: http://ow.ly/zXYvh #IBMDataMag
"Transformative Analytics" (http://ow.ly/zXY3A ) JK--IBM Press book: discover how IBM realizes value frm big data & analytics #IBMDataMag
RT @kexpplaylist #kexp Bad Betty by The Sonics JK--2014. Believe or not, new single from the legendary 1960s
Seattle rockers. Very
"Establishing Data Governance for Enterprise Business Terms" (http://ow.ly/zXW7Y ) JK--Case study: Canadian National Railway #IBMDataMag
"Broadening Support for Apache #Spark " (http://ow.ly/zXVKr ) JK-- #IBM expanding commitmt for #Spark large-scale proc engine #IBMDataMag
"Hadoop Dev for Developers by Developrs" (http://ow.ly/zXViO ) JK--Visit Hadoop Dev site 4 info on Hdp/InfoSphere BigInsights #IBMDataMag
RT @kexpplaylist #kexp My Sister by The Juliana Hatfield Three frm Become Wot U R JK--1993. One of my fave 90s rocksongs. Real girl feelings
"Circumvent IT Project Backlog for Effcient Ops" (http://ow.ly/zXUnr ) JK--Aartika Sardana on data mgt in app consol/retiremt #IBMDataMag
"Automate Conversion to Native SQL Procedures" (http://ow.ly/zXTWU ) JK--Nate Murphy on implementng DB2 native SQL procedures #IBMDataMag
"Virtualize Nonproduction Testing Ecosystems" (http://ow.ly/zXThq ) JK--RHughes & HSuri on #IBM Netezza Platf Devel Software #IBMDataMag
Geospatial analytics? Ammunition against pestilence (http://ow.ly/zXOiZ ) Tuesday #IBMDataMag quick-hit
Book on "1001 Movies You Must See Before You Die." Viewing all of those will consume the rest of my natural lifespan.
"Ask 'crunchy' #bigdata questions tht lead 2 breakthru analytics" (http://ow.ly/zWbEQ ) JK--Defines as Qs alignd w/burning biz imperativs
"History o Ensemble Methods" (http://ow.ly/zWaJY ) JK--"Build large # o somewot indep pred modls, combin by vote/avg 2 yield vry hi perf"
OpenML (http://ow.ly/zWaf7 ). Yet another new open data-science community. Collaborative development of machine learning models.
Most IT indstry "infographics" have all th clean, crisp focus of a Rube Goldberg machine that's been rendered graphically by Jackson Pollock
When faulty-phone-connection crosstalk content is more interesting than the audioconference you're on, that's a problem of a different order
"Role of NLP in #BigData" (http://ow.ly/zW7To ) JK--Good. Discussions of oil & gas industry apps of natural language processing are rare
"The Geography of Wikipedia Articles" (http://ow.ly/zW7ty ) JK--Wow. Needs to be correlated to individual contributors.
"Google Translate's Lead Scientist...Help Conquer Death" (http://ow.ly/zW1MH ) JK--Strategic overreach. Grim Reaper already knows endgame
First impressions are important as framing devices. I think of them in a cinematic sense as the "establishing shot."
Another one of those online data-scientist professional-services marketplaces: Experfy (http://ow.ly/zVzOs ). Interesting.
#BigData on the move? Emergence of the mobile back-end as a service p1: http://ow.ly/zVqSL p2: http://ow.ly/zVqV4 Mon #IBMDataMag qh
Drafted latest #IBM jk #InfoWorld column: "Monetizing your existence as a crowdsourced data scientist"
RT @kexpplaylist #kexp The Wind by Cat Stevens from Tea for the Tillerman JK--1970. His best work had devotional soul at its very heart.
New @IBMSmtWorkforce offerings combine #social, #cloud and #analytics to add precision to talent management http://ibm.biz/BdF4gx
"Blistering fast SQL access to your #Hadoop data" (http://ow.ly/zV6eU ) JK--Fatma Ozcan on #IBM #BigSQL 3.0 performance benchmarks.
A lot of things on the Web are horrible timewasters that you can't really know are a waste of time till you've wasted the time.
Dumbest come-on in online publshng that's probably effective eyeball lure: "[such&such] could change yr life." If U click & it doesnt, tough
RT @kexpplaylist #kexp Femme Fatale by Velvet Underground from VU+
N JK--1967. I luv
th faint flat almost comical "she's a FF" chorus on this
Actress Cheryl Hines weds RFKJr. Couple intro'd by Larry David. Would be gr8 "Curb Yr Enthusiasm" plot premise. Where would Larry take this?
Daughter & I did a great speedwalk last night. Actually, didn't start as speedwalk, but we hustled halfway due to impending electrical storm
Toxic algae bloom in western Lake Erie contaminates
Toledo water supply. Industry not the prime
culprit. Agricultural runoff has run amok.
In border-control news,
governor Rick Perry says we need to keep out Syrian terrorists by fortifying
the Mexican border.
At one point or another, almost every member o my extended family has had email hacked. When siblings & cousins call me "James," it's unh-uh
Did Whole Foods and Trader Joe's in a single shopping trip. Have no use for the former, love the latter. Dont need overpriced frou-frou food
Listened to this Bill Gates interview from Comdex 1986 (http://ow.ly/zTdpl ). Hard to believe that's almost 30 years ago. Ah, nostalgia!
Dining with the girls at Tippy's
Unpretentious little joint in Tacos
Brookfield Plaza South Sprinfield
with good enchiladas.