Friday, April 04, 2014

Aweekstweets March 28 to April 4 2014: the week that was riddled with randy randomness

Nice thing about offering certain people a chance to contrib to #IBMDataMag is that it gives me the opportunity to let em know I'm their fan

 16m
Catch me speak @ Collaborate, Vegas, this Tues "#IBM Data Warehousing: Modern Ecosystems Designed 4 #BigData Analytics" #oaug #C14LV #ibmblu

 18m
My only tech-related tweets today were of the things I'd personally authored. Not a lot of interesting stuff in my reading today. Yawn!

 32m
New #IBM jk blogpost: "#IBM #bigdata roundtable: The transformative power of big data" (http://ow.ly/vscEU )

 8h
Akella: Attribution modeling. With 360-view & multi-screen M&E experience, "how do you know what people are responding to?" #bigdatabytes.

 8h
Akella: "Statistical munging can make for pretty good privacy while giving [consumers] what they want." #bigdatabytes

 8h
Akella: Mobility changes things. "Do people like diff types of things [media & entertainment] in big screen or small screen?" #bigdatabytes

 8h
Ram Akella, prof IS & Tech Mgmt, UC-Santa Cruz discussing converge o #bigdata, #datascience, & economics in online advertising #bigdatabytes

 8h
@graemeknows leading the video panel on #bigdatabytes. Never mind the tech difficulties. Graeme can work miracles. It'll happen!

 8h
Waiting for #bigdatabytes to start. Live videochat "Zooming In: Seeing True Value of Audiences .” Watch http://ibm.co/BDBytes 

 8h
US Postal Service is now primarily a carrier of promotions, prospectuses, and privacy statements. Dump dump dump dump dump!!!!

 12h
I try not to follow tweeters whose only tweets are simply announcements that their "daily is out." I'd prefer those dailies stay in closet.

 13h
RT @kexpplaylist #kexp Dead Is Dead by Terry Lee Hale from Sub Pop 200 JK--1988. Acoustic piece slightly predates early Seattle grunge scene

 13h
Whew! Drafted all 5 nxt-wk qhs. Even quoted something I blogged 9 years ago at length (http://ow.ly/vqx6F ), updated 4 IoT. Still applies

 16h

 16h
Moneyball? p1: http://ow.ly/vqavw  p2: http://ow.ly/vqawY  Friday #IBM quick-hit

Letterman retiring. Love his quip: "What this means now is that Paul and I can be married."

I have no opinion on the Dems' DC mayoral nominee. But I do know I like saying "Bowser." Wasn't that the lead singer of Sha Na Na?

Streaming vid playback of today's #IBM #BigData Roundtable that I moderated: http://ow.ly/vpe8j . Thanks to all panelists & participants!

When I glance at text, I look for headers, lists, tables, &/or parallel constructions: clues to author thought INner STRUCTure. In-STRUCT me

"#BigData showdown: Cassandra vs. HBase" (http://ow.ly/vp9Dr ) JK--Good comparison, but not a "showdown." Each has distinct use cases.

"Get inside data scientists' minds through touchy-feely talk sessions" (http://ow.ly/vp9lo ) JK--OK, but keep your hands to yourself.

"Teaching an Old Log New Tricks w/ Machine Learning" (http://ow.ly/vp99v ) JK--Good piece from Accenture. Wish I'd thought of that title

"No one thought about it in America: parasite that uses yr body as personal feeding ground." Huh? We DO understand the concept of parasites

Redirecting...if you are not automatically redirected, please power off your devices and redirect yourself to some place outdoors. Breathe!

If you accept that data scientist is sexiest job around, what do you ask a job candidate that won't get you sent to sensitivity training?

Whew! Glad that's done. Finished moder8ng hourlong #IBM video spreecast on transformativ #bigdata. Waited till last 5 mins to insert own POV

Usual day: trying 2 sustain distractible focus on 1 topic @ time in order 2 draft 1 crystal POV after another on shifting mix o complex crap

New #IBM jk blog: "#BigData powers the practical know-it-all in us all" (http://ow.ly/vnTHA )

Prediction markets? Fostering open marketplaces for models and modelers p1: http://ow.ly/vnQmr  p2: http://ow.ly/vnQoG  Thurs #IBM q-h

A8: Look for diverse thinkers & ppl willing to try something new – but who have a sense of pragmatism #bigdatamgmt
 Retweeted by jameskobielus

RT @mt_marko A8 Think big, start small and try to produce value at early stages. Don't be afraid of failures, retry. #bigdatamgmt
 Retweeted by jameskobielus

A8: Hire a scientist who is versed in research methods for #bigdata team. #bigdatamgmt
 Retweeted by jameskobielus

A8: #bigdatamgmt Establish #bigdata center of excellence to drive ongoing cross-org team recruitment, training, & skills/knowledge sharing

A8: #bigdatamgmt Build a #bigdata team that is prepared to productionize all of their work 24x7 in biz operations with stringent SLAs.

A8: #bigdatamgmt #BigData continues to evolve in every direction. Make sure your teaming strategy is agile over time to keep up.

A7: outsource first then aum for mid-term independence: deploy in house big data #IaaS within 5 years #bigdatamgmt
 Retweeted by jameskobielus

A7: Domain expertise is tough to outsource. Adequate business context is always a challenge for consultants. #BigDataMgmt
 Retweeted by jameskobielus

A7: #bigdatamgmt If outsourcing because need to address surge requirements, don't do permanent hires if you don't expect the surge to last

A7: #bigdatamgmt If new to #bigdata & need 2 bootstrap own capabs, outsource core tech/dev functions but with plan to in-source over time

A7: #bigdatamgmt Key consideration: DON'T outsource core #bigdata program mgmt or domain expertise. Retain those 4 biz align & execution

There are at least 12 new "Chief" roles being evangelized today. @jameskobielus: A6: #bigdatamgmt The words between "chief" & "officer".
 Retweeted by jameskobielus

A6: #bigdatamgmt I think "Chief Analytics Officer" gets to heart o what so-called #CDO manages, per what drives differentiated outcomes.

@BlairReeves #bigdatamgmt No. Function might B shared across entire enterprise (e.g., multichannel engagemt), in which case serves many LOBs

A6: #bigdatamgmt The word between "chief" & "officer" could be the resource, entity, or outcome managed. I think "data" too limiting.

A6: #bigdatamgmt U need C-level champion of data/anaytics driving biz outcomes. Doesnt matter what that job's called: #CDO #CIO etc

A5: Someone who ustands #analytics in all its glory (stats, machine learning) might be a good leader for a #bigdata team #bigdatamgmt
 Retweeted by jameskobielus

A5: #bigdatamgmt If, say, doing social sent analysis, biz lead is CMO, dev lead maybe data-scis, deploy lead by #Hadoop et al specialists

A5: #bigdatamgmt C-level should lead with vision, charter, & funding. IT proj mgmt pros should lead team recruitment, training, & tasking

A4: #bigdatamgmt I'd NOT include someone if they DON'T have DevOps orientation, or a needed IT specialty, or biz domain knowledge.

A4: #bigdatamgmt I'd built a team where everybody has a common DevOps orientation, includes chief tech specialties, & has domain experts.

A4: #bigdatamgmt Wouldn't ostracize "types of people" from team. But would look for candidates well-versed in the techs & the apps

A3: In digital analytics, I’ve seen a hub-and-spoke model have the best results #bigdatamgmt
 Retweeted by jameskobielus

A3: Center of Excellence w/ reps from diff groups is a good model, says @jameskobielus, http://ibm.co/1aqq7rf  #bigdatamgmt
 Retweeted by jameskobielus

A3: #bigdatamgmt If each line-of-business has distinct #bigdata plats/apps/requirements, may make sense to distribute teams among LOBs

A3: #bigdatamgmt If you have a multi-zone arch, U may need distinct teams managing in-memory marts, RDBMS EDWs, & #Hadoop landing zones

A3: #bigdatamgmt #Bigdata team should be confined to small locked room at all times. Or not. Align team(s) w/key #bigdata plats/apps/domain

A2: For a #bigdata team, you’ll need a program owner, technical experts, business analysts, and project manager. #bigdatamgmt
 Retweeted by jameskobielus

A2:Evangelist who drives the change in your org to became data driven org is essential #bigdatamgmt
 Retweeted by jameskobielus

A2: #bigdatamgmt Where data integration skills are concerned, need emphases on unstructured ETL, stream computing, and data governance.

A2: #bigdatamgmt Where stat modelers come in, need data/content mining, predictive modeling, graph modeling, & sentiment analysis

A2: #bigdatamgmt Where data scientists & app devs are concerned, need blend of statistical modelers, domain experts, and programmers

A2: #bigdatamgmt Key roles on #bigdata team include data integ specialists, data platform admins, data scientists, & app developers.

A1: #bigdatamgmt What you should also identify first is how to grow your team & tech to support #bigdata target architecture/vision

A1: #bigdatamgmt What U "select" first for your #bigdata proj is biz strategy/use cases to be addressed. Team & tech selected to fit that

A1: #bigdatamgmt Often, org already has "yolk" of #bigdata team "egg": DBAs, stat modelers, etc, plus core of tech plat (eg EDW) to evolve

A1: #bigdatamgmt It's a chicken-and-egg issue. You don't need to choose one first. As you build your team, you're also assessing techs.

@IBMbigdata @analyticshero Happy to be tweeting about teaming on today's #BigDataMgmt

@clivel_98 To oversee "deadwood staff"? How we dehumanize people's jobs says a lot about whether we respect the value they add, or otherwise

Join @analyticshero, @thomasdeutsch, & li'l ol' me for #BigDataMgmt chat on "Building a #BigData Team." Today. 12 ET! http://ow.ly/vkDXO 

"Defining the Chief Data Officer" (http://ow.ly/vkxHT ) JK--Infographic describes 3 models for CDO organization. #IBMDataMag

"Warm Sun and a Hot Agenda" (http://ow.ly/vkxeN ) JK--Nancy Hensley reports from 2014 IIUG Informix Conference in Miami. #IBMDataMag

"Recapping 2014: Significant Trends for #BigData" (http://ow.ly/vkwQz ) JK--Tom Deutsch on privacy, Xaas, & open source #IBMDataMag

"Taking Swift Action to Migrate and Consolidate Data" (http://ow.ly/vkwvO ) JK--Aartika Sardana on #IBM IIS for SAP Apps. #IBMDataMag

"Igniting the Future of Cognitive Computing" (http://ow.ly/vkw0n ) JK--Alex Philp on #Watson spatial analysis apps in IoT #IBMDataMag

"Daily Routines o Geniuses" (http://ow.ly/vknkl ) JK--Metrix. Supprtiv othr. Distract- busywrk- socializ-averse. Dont stop on stuck. Walk

"How Rise Of "R" Computer Language Is Bringing Open Source 2 Science" (http://ow.ly/vkn2u ) JK--Every computational scientist will use R

"Japan nuclear disaster could strike in US" (http://ow.ly/vkmbw ) JK--2011. Wha...? @RitaJKing also a nuclear expert? Renaissance woman!

"Probabilistic Models of Cognition" (http://ow.ly/vkl1z ) JK--Regular reasoning in a world riddled with randy randomness

"How Corporate Departments Totally Mishandle #BigData" (http://ow.ly/vkkFf ) JK-- #IBM @JeffJonas on "outdated info-sharing policies"

Not too shabby. Just shabby enough. Life without shabby? Unthinkable!

"Analytics = Manipulation of Data Structure" (http://ow.ly/vkkcj ) JK--Nicely ties data modeling to enterprise architectural modeling

I live in the 'burbs, but I do have an opinion in the DC mayoral race. I'm glad Gray was dumped by Dem voters yesterday. What a turd.

2 years ago today, I was in Austin 4 #IBM new-employee orient. 32 years ago I'd've gone 2 UT-Austin if UW-Madison hadnt made me better offer

Peta-governance? Where trustworthiness concerned, proof is in data-gov process p1: http://ow.ly/vkcrG  http://ow.ly/vkcsv  Wed #IBM qh

LinkedIn. The social where people don't socialize. They just connect, check each other out, endorse each other liberally, and pitch resumes.

Breaking news. New debris spotted. Old debris has drifted. Some vaguely disgusting debris surfacing from the deep. We'll keep you posted.

Photogenic/telegenic. Professional words re person whose physical appearance is presentation asset. Avoid saying "beautiful"/"handsome."

"IIC Forms to..." (http://ow.ly/vjjWf ) JK--Mandate includes "building confidence around new and innovative approaches to [IoT] security"

IBM has transformed itself into a Big Data and Analytics powerhouse." More in @_ChrisVersace's Q&A w/ @BobPicciano: http://onforb.es/1llfvMV 

Join #IBM chat tomorrow 12-1pm (EDT) on "Building a #BigData Team" w/ @analyticshero myself et al http://bit.ly/1jXMchZ  Follow #bigdatamgmt

"Cross-industry IoT group pushes for gear that works together" (http://ow.ly/viQ3A ) JK--#IBM, AT&T, Cisco, GE, and Intel

"Review: HBase is massively scalable -- and hugely complex" (http://ow.ly/viPPj ) JK--Good analysis o #Hadoop #bigdata platform component

When somebody sez somebody a "rock star" these days, not usually referring 2 rock musicians--& few new mega-music-stars anymore are rockers

"#BigData, Synthetic Biology & Space Planes Are Weapons of Future" (http://ow.ly/viNX0 ) JK--About US DoD DARPA's "Big Mechanism" program

"Apache Falcon-Data Governance 4 #Hadoop" (http://ow.ly/viMUl ) JK--Data gov engine auto-gener8s data pipeline workflows in Apache Oozie

"Apache #Mahout, #Hadoop original machine learning proj, is moving on from #MapReduce" (http://ow.ly/viMxw ) JK--Being reworked 4 #Spark

Hey LinkedIn, when suggesting a post for me to read, try not suggesting my own. Point me at something I dont have banging around in my brain

"How #BigData Shapes AT&T's Adv Creative" (http://ow.ly/viKaY ) JK--"It's not complicated" campaign. Appeal isn't: cute kids, cute lady

Hmm...LinkedIn thinks today's my 2-year #IBM anniversary. It's actually tomorrow. Not a blue-newbie anymore. I'm deeply blue. (M) Go Blue!

LOL when some1 misspells easy English word w/insertd letters on billboard. Saw "teh" (4 "the"). How, standing there, did installer not see?

"Haboob Clouds #Hadoop Future" (http://ow.ly/viEA8 ) JK--Likens latter to former (Arabic word for dense dust storm).

"From the trenches: 360-degree data science" (http://ow.ly/viDM1 ) JK--VGranville assesses his own data-science site's effectiveness

"Top 10 Capabilities for Exploring Complex Relationships in Data for Scientific Discovery" (http://ow.ly/viDjz ) JK--Gr8 @KirkDBorne blog

New #IBM jk #ITKE column: "Extract maximum insights before you purge your data" (http://ow.ly/vivmz )

Data journalism? p1: http://ow.ly/vhyVC  p2: http://ow.ly/vhyXV  Tuesday #IBM quick-hit

Love Esquire's list of things a man should do before he dies. "#85: Get handsomer."

My ideation flow: readings gener8 idea tweets, which are nuclei o specific quick-hits, which spawn blogs, which crystallize my evolving POV

A8: #CXO Go IoT big-time. Truly #digital customer of 2024 will rely far less on handheld gadgets & far more on connected homes, cars, etc.

A8: #CXO Stay gadget-agnostic. Truly #digital customer of 2024 won't use any devices they have in 2014. 100% gadget churn every few years.

A7: #CXO Biz that succeed are those who leverage social 4 engagement, cloud 4 agility, mobile 4 access, & IoT 4 sensing/optimization

A7: #CXO In today's world, the biz imperative is don't fool yourself into thinking old rules apply. New techs disrupt at astonishing speed

A7: #CXO First step is 4 biz 2 develop vision of evolving customer-centric landscape of 2020. Second step is move in that direction ASAP

A6:Arm your staff with mobile #AR apps that identifies C, knows his profile/history/intents & gives advices how to interact with C #cxo

A6: #CXO Biz can empower employees by tailoring channel conversations through embedded VoC, NPS, social influence, & other key analytics.

A6: #CXO Business can empower employees to engage smartly 24x7 by embedding customer analytics into every channel-facing app.

A6: #CXO Businesses can equip employees for customer-driven world by giving them full mobile access to 360-degree cust view at all times.

A5: Allowing customers room to breathe with less digital bombardment..they need to have a good night sleep to make better decisions #cxo
 Retweeted by jameskobielus

A5: #CXO Being "too connected to customers" might mean the biz loses sight of need to diversify to other customers.

A5: #CXO In terms of #loyalty, "too connected to customers" might be cust retention rates so high U risk taking 4 granted. Woo them anyway!

A5: #CXO Biz being "too connected to cust" is how might describe creepier aspects o intrusiv mktg enabled by #bigdata. Observe boundaries!

A4. Big Data fills in the gaps by providing closer to 360 degree insights into customer behaviours, than smaller subsets of data. #cxo
 Retweeted by jameskobielus

A4: #cxo Bigdata can help glean actionable Attributes at a more granular level leading to more conversions...and retain loyalty

RT @thecxguy A4: Most often, data help us to quantify intuition and validate instincts to convince decision-makers of best course. #cxo

A4: #CXO High-perf, efficient, consolidated customer #bigdata platform helps you control costs as your multi-channel environment expands

A4: #CXO #BigData is pivot point where all your customer-facing channels converge on a unifying principle: 360-degree 24x7 engagement

A4: #CXO #BigData provides the deepening historical record of customer engagements & drives a more nuanced picture of how it should evolve.

A3f: Social media enable us to understand in greater detail the impact our actions are having in real-time #CXO

A3: First have to find the conversations, and then engage customers on their preferred channels. #CXO

A3: #CXO Businesses create sustainable customer engagement by focus on differentiators: price, value, service, community.

A3: #CXO Customers are not "firmly" in the driver's seat. They are influenced by 360-degree brand experience the biz fosters. Dazzle them.

A2. You can build predictive models, to predict customer intention, but so many behavioural variables that can throw things off. #CXO
 Retweeted by jameskobielus

A2: #CXO Biz can confidently predict cust actions by recognizing repeated, patterned nature of many behaviors. We're consumers of habit

@editingwhiz A draft is something other than a final published piece.

A2: #CXO Best "get there before" strategy is data-driven cust profile compiled from interaction history & tuned via predictive analytics
 Retweeted by jameskobielus

A2: #CXO Biz influences the customer's next action(s). "Getting there before the customer" means anticipating how they'll respond to you

A2: #CXO A business can CONFIDENTLY predict customer next action, but "accurately" (ie, perfectly) is too optimistic. Even customers can't

A1: Biz need 2 be aware that 81% of cust reach out on social networking sites 4 advice b4 purchasing prod < listen there #cxo
 Retweeted by jameskobielus

RT @rgmarkey: A1d: Don't be satisfied w/ mere demographics (or, for B2B, "firmographics"). How do they REALLY buy? Speak w/ individuals #CXO

A1g: Don't forget that your salesforce, store employees, call center reps and other front line employees offer tremendous insight #CXO
 Retweeted by jameskobielus

A1: #CXO Understanding how customers buy TODAY means gathering & analyzing every new day of customer data afresh. Needs change daily

A1: #CXO Really understanding the customer requires depth what they're NOT telling you. What's pent-up in their hearts? They angry at you?

A1: #CXO Any biz hoping to REALLY understand the customer must walk in their shoes. Map their buying journey. Is it rough road? Smoothe it

A1: #CXO Customers buy things. They also DON'T buy things. They can choose to put you out of business. Don't take them for granted.

@IBMbigdata Hi, hello, howdy, hi-dey-hee, & hi-dee-ho. Glad to be among you today tweeting on #CXO about thriving in a customer-driven world

@editingwhiz People draft beer too. Verbs can have many senses.

CountDwn 2 #CXO: Thriving in a customer-driven world w/ @rgmarkey @davemitz y yo http://bit.ly/1jEa3ra  CU in 90 minutes! #custserv #cxpa

Drafted latest #IBM #AnalyzingMedia column: "Sports Teams Haven't Even Begun to Smack That Ol' Moneyball Right Out of the Park"

Drafted latest #IBM #Dataversity column: "Don't Understaff And Overstretch Your Analytics Development Team"

Drafted new #IBM #ITKE column: "Extract maximum insights before you purge your data"

Drafted new #IBM #InfoWorld column: "Resist the new-age push to an all-in-one database basket"

Drafted my next #IBM blog: "#BigData powers the practical know-it-all in us all"

We sure got a good soaking over the weekend. Walking in my super-saturated yard to investigate any storm damage. Squish squish squish squish

We had a great discussion about #IBMBLU on #Cloud.... did you miss it? Conclusion? DW is perfect for the Cloud: http://bit.ly/1dBm4LC 

Healthcare analytics? Using advanced image analytics to spot hidden cancer patterns (http://ow.ly/vd85O ) Monday #IBM quick-hit

It's nice when somebody "like"s one of my Linked posts from months ago. I then re-read to determine whether I still like it.

Ah, th romance of sleet! What I wouldn't give for a wintry mix.Yikes, that's not snow is it? Sorry about tempting U God. Can I take it back?

RT @kexpplaylist Driving Me Mad by Neil Finn from Live at KEXP JK--Totally digging this live-right-now. Finn's still the perfect songster.

RT @kexpplaylist #kexp When The Night Comes Falling From The Sky by Lucius JK--Another great one from this group. Dylan cover, apparently

Read article about London's skyscraper-construction boom. Saw artist depiction of emerging skyline. I'm impressed. 21st Century world city.

Forced myself to watch new SNL last night. Weak writing & undistinctive cast. Prefer Armisen's Obama to Pharaoh's. I dozed off. Didn't laugh

Where's the damn lamb this month was supposed to go out like? The lamb hasn't shown his head. Not even once. Nothing even remotely lambly!

Lucius "Turn It Around" (http://ow.ly/vamZP ) JK--This gr8 song/video playd @ my fitness club th othr day. I looked around: "#KEXP song?"

I do follow some tweeters who occasionally tweet in languages other than English. But if their % of English tweets falls below 75%, I don't

"7 Things Remarkably Happy People Do Often" (http://ow.ly/vajVH ) JK--Perusing lists like this = gr8 way 2 make yrself remarkably unhappy

Doing my usual mental de-dupe of the tech headlines that these e-newsletters recycle over and over within a 1-2 week "it's still news" cycle

RIP Adm Jeremiah Denton Jr. Incredible bravery. Vid o him blinkin Morse Code 'T-O-R-T-U-R-E' as POW in North Vietnam...

College athletes can unionize? What are they going to demand at the negotiating table? Allow keg parties to earn credits toward graduation?

ACA signups top key target. But hasn't attained true goals: strangling liberty, tanking economy, & making Obama dictator. Working on those

Speedreading is the best (and fastest) way of identifying well-written material. The structure, analysis, & thesis jump right out at you.

I like current Dilbert story arc about the constant-traveling consultant with the unhealthy lifestyle. Living out of a suitcase is hazardous

Does CNN even cover other news anymore? Crew incapacitated? Seem 2 B autopiloting newscraft into ocean of oversaturated 1-story overkill