Friday, January 31, 2014

Aweekstweets January 24-31 2014: the week Pete Seeger went to flowers

US patent #8,375,213: Systems and methods for enabling trust in a federated collaboration ( ). Issued 2-12-2013

Re-reading one's own work after an intense writing session is like awakening from dream. Was that train running down the track in my head?

Finished drafting all o next week's quick-hits this morning. Re-reading. Connected a lot of conceptual dots across my readings & my writings

"Deep Web Intelligence Platform: 6+ Capabilities Necessary 4 Finding Signals in Noise" ( ) JK--Automated sensemaking

Today's "only Seattle artists" day on #KEXP reminds what a stunning diversity of hi-quality music continues to flow from the Emerald City.

RT @kexpplaylist #kexp Lion's Mouth by Arthur & Yu JK--"Oh my fingers in your buttons are like kissing cousins making fabrics come undone."

"#BigData’s Dangerous New Era of Discrimination" ( ) JK--Michael Schrage on the dark side of demographic segmentation

"The Data Storm: Retail & The #BigData Revolution" ( ) JK--Interactive report from The Economist's Intelligence Unit.

"Fundamentals o Good Data Narration" ( ) JK--Eric Swayne: "Critical not to confuse the visualization WITH the insight."

"Why data science matters to Foursquare" ( ) JK--Location-contextual notifications boost user time on check-in app.

"#BigData in Education: Big Potential or Big Mistake? ( ) JK--Potential: personalization, feedback, tracking, assessmt

"#Bigdata ... fast data w/BlueDBM" ( ) JK--MIT prototype multi-node DB, networked flash devices, "near real-time query"

"Was Eliot Ness a Hero or Hollywood-Inspired Myth?" ( ) JK--Chicagoans believe former. Mrs. O'Leary's Cow got a bad rap

Sexy statistics? The vintage kick of old data poured into fresh analytic bottles ( ) Friday #IBM quick-hit

Stupid WSJ headline says retailer mailing to inadvertently offensive address is fault of "big data," when in fact scale of DB is irrelevant

Wife said she mistakenly mispronouncd happy new year in Hakka & came out dirty. Have 2 trust her there. She & another Chinese had good laugh

It's tomorrow now in Asia, so I'll wish Tio Ciu family & friends (phonetically) "sin cha ju i." Happy New Year. Horse.

"The Data Visualisation Catalogue" ( ) JK--Cool compendium of alternative ways of eyeballing & exploring data patterns

I'm a bit jaded about "year of" IT pronouncements. Seems2me, the blogger usually means "year of [whatever I personally happen 2 B doing]."

"Use AR 2 cleanse, analyse, & represent #BigData? [p1]" ( ) JK--3-D visual vetting in real-world geospatial context

"Hal Varian & “New” Predictive Techniques" ( ) JK--"Techs that R more well-known in ML than econometrics or std stats"

"The Two Characteristics of Data Accuracy" ( ) JK--1) Right. 2) Unambiguous. Also. well-formed & consistent format.

@mjcavaretta Looks good. Thanks for the great automotive #bigdatamgmt tweetchat yesterday, Mike!

"Will #Hadoop replace traditional #DW?" ( ) JK--Ford's @mjcavaretta quotes something I blogged 3 years ago. Still valid

New #IBM jk #Infoworld column: "Cognitive computing can take the semantic Web to the next level" ( )

Recommendation engines? Untapped potential of video, image, & gesture analytics in retail showrooms ( ) Thurs #IBM q-h

New #IBM jk blog: "Autonomic planet: Distributed intelligence for a self-healing ecosystem" ( )

Preaching to the choir is no sin. It's only a sin if you fail to preach to them about the gospel of harmonizing in the right key.

#bigdatamgmt @SJAbbott @marksalke @ Out of deference to the tweeter from Ford, I opted not to mention their rival: Porter. 1928 model year.

Did intense 2-hour analyst call taking open-ended ad-hoc questions on complex #IBM response. So gr8ly appreciate having my colleagues on it!

"5 ways Washington DC is very different than Silicon Valley" ( ) JK--Also, we tend to wear suits.

A8: #bigdatamgmt My car GPS talks to me. It provides endless merriment when mispronounces common words. Moves me deep down in my funnybone

A8: #bigdatamgmt "Talking cars"? Is this like a vehicular Mr. Ed? You mean connected cars?

A7: #bigdatamgmt Most of the #auto #telematics car chatter shouldn't be stored long-term. Most is disposable. Needs to be purged often.

A7: #bigdatamgmt I'd prefer that my in-car #telematics chatter doesn't store the choice expletives I occasionally direct at nearby drivers

A7: data car gener can be divided 2 different classes based on relevance for manuf, dealers, societal, privat, insur... #BigDataMgmt
 Retweeted by jameskobielus

A7: #bigdatamgmt Owners should store/control #telematics chatter assoc with in-car entertainment. Private/monetizable personal asset

A7: #bigdatamgmt Depending on laws, #auto manufacturers should have access to warranty-related #telematics data while in effect.

A7: #bigdatamgmt Will store #auto #telematics "car chatter" in owner's cloud. Ideally, tags will identify data assoc w/each licensed driver

A6: #bigdatamgmt Absolutely DO NOT focus on alerting drivers through smartphone or other gadget apps. Far too distracting. Dangerous

A6: #bigdatamgmt Make full use of the in-car navigation system to GUIDE drivers in real-time (e.g, detour here). Don't JUST alert them.

A6: #bigdatamgmt Use heads-up displays, audio, maps, graphics, large fonts, & even tactile/kinesthetic approaches to send alerts

A6: #bigdatamgmt Regard driver as "functionally disabled"--ie=., eyes, hands, & attention on safe driving. Dont distract w/excessive alerts

A5: #bigdatamgmt Component performance data should B checkpointd in factory & upon delivery. What's design/engineering/manuf's fault?

A5: #bigdatamgmt Key to correl8n is timestamped + geostamped #telematics event data. Fine-grained pinpointing o failure-contributory events

RT @Natasha_D_G A5: Geospatial, time series and machine data analysis is required. #bigdatamgmt #telematics

A5: #bigdatamgmt If you have road/envir/operation sensors in cars, you have complete driving history available to correl8 w/maintenance

A4: #bigdatamgmt #Telematics data can reveal if drivers fail to add oil & coolant, hence stressing engines beyond manufacturer recomm.

A4: #bigdatamgmt #Telematics data can reveal if drivers fail to avoid potholes, hence throw out their alignment & damage undercarriage

A3: #bigdatamgmt Info from vehicles & OLTP systems should be combined with sentiment, #CXO experience, & survey data from owner/drivers

A3: #bigdatamgmt Building & managing 360-degree car "view" requires #InternetofThings + #cloud + #BigData.

A3: #bigdatamgmt Combining event data-in-motion (from in-car #telematics) with at-rest data data requires streaming/Hadoop/NoSQL integr8n

A3: #bigdatamgmt All that car-focused data will be combined, per privacy constraints, by #auto owners, insurers, & dealers.

A2: #bigdatamgmt I expect that autonomous vehicles will be deployed everywhere as continuous road-quality sensor/reporters.

A2: Think in terms of collapsing time & accuracy issues between event and those who did the design / engineering in the event #bigdatamgmt
 Retweeted by jameskobielus

A2: #bigdatamgmt If you have 1000s of cars/drivers feeding road-quality data continuously, provides richer data for diagnosing faster

@mjcavaretta A2: #bigdatamgmt For consumer-owned car, sensors already on vehicles. For traffic-mgt/civil-engineer cars, specialized sensors

A2: Vehicle sensor data w/ warranty claims, production data & other data can ID quality patterns faster than b4 #bigdatamgmt
 Retweeted by jameskobielus

A2: #bigdatamgmt Cars are the advance scouts of any issues in and around the road system. They & their drivers see/feel problems first.

A1: #bigdatamgmt Road-embedded sensors can alert engineers to emerging issues (e.g., bridge stress fractures) before catastrophe hits.

A1: #bigdatamgmt In-car #telematics can provide real-time intel to feed predictions of hazard systemwide impacts (e.g., congestion)

A1: networked traffic signals could get info from vehicles & reroute traffic or change signals for better flow #bigdatamgmt

A1: Lightweight MQTT transport protocol enables real time communications #bigdatamgmt #auto #telematics
 Retweeted by jameskobielus

A1: Combo streaming geospacial data + vehicle sensor data extends driver's awareness beyond range of vision #bigdatamgmt

A1: #bigdatamgmt #Telematics can use in-car imaging, motion sensors, environmental sensors to detect and assess road hazard in real-time

A1: #bigdatamgmt Connected cars should be networked in real-time to regional alerting service. First car on scene alerts the system

@mjcavaretta I stand corrected!

#bigdatamgmt I am deeply honored to be doing this tweetchat. And I hope you all feel honored to have me in your virtual presence.

Special treat on today's #bigdatamgmt tweetchat: 2 Livonia geeks. Myself + Ford's @mjcavaretta. I'll forgive him for being Bentley HS alum.

@Natasha_D_G knows I have her back on the #bigdatamgmt tweetchats. Have topic, will tweet.

40 min to #bigdatamgmt chat: Automated #Telematics: When Cars Talk w/ @mjcavaretta  Join me/us! #auto #bigdata #hadoop

"Social Sentiment’s Missing Measures" ( ) JK--Classif by "biz-aligned splits." Sentiment density. Variation. Volatilty

Internet of Things? Instrument the birdies, bees, and other beasties p1:  p2:  Wed #IBM q-h

Join me et al on #bigdatamgmt chat: Automated #Telematics: When Cars Talk  at 12noon ET today #bigdata #auto

RT @kexpplaylist #kexp Desire Lines by Deerhunter from Halcyon Digest JK--2010. I pound a tweet every time I hear this pounding beat.

Catch me downtown tonight, "Past, Present, Future of #DataScience Education," GWU, WashDC, Funger Hall, Rm 103. One of many Meetup attendees

WashPost article on a "culture of toxic military leaders." I'm sure that career encourages nasty power-besotted SOBs to indulge their worst.

Day greeted me with an inch. I greeted it with a shovel.

Saw "Blue Jasmine" on DVD. Wow! My fave Woody Allen & fave Cate Blanchett film of all. Superb script, direction, & performance.

Fox News respectfully notes Seeger passing. Paraphrasing: "he sang catchy tunes." Roughly what you'd say if Billy Joel'd bit the big one.

"Large scale data analysis made easier with SparkR" ( ) JK--Blog from UC-Berkeley AmpLab

"#BigData Econ" ( ) JK--Knwldg w/out expertise, scale w/out mass, data = capital, privacy is brand value, semantic econ

"Eclipse Ups Ante for Internet of Things Community" ( ) JK--Now has 13 projects focused on open-source M2M & IoT apps

@GrandpaRobot You seem to have ideas on the topic. Why not write that blog yourself? I'd love to read it.

"#IBM Anncs Glob Cnsltg Pract" ( ) JK--IBM Interactive Exper. Life Event Detection. Behav Pricing. Psycholing Analytics

RT @kexpplaylist #kexp Talking Union by Pete Seeger frm If I Had A Hammer: Sngs o Hope+Struggl JK--1947. Written 1941 by Seeger/Lampell/Hays

Weekly process of forcing my head to revisit data mgt themes I may not have blogged on in a while. Data gov, for example. What's new to say?

"New #IBM Kenexa Talent Suite Taps #BigData" ( ) JK--HR analyz empl work exper, social enggmt, skill dev, indiv intrsts

"Streets o Ann Arbor Will Soon Be Filled w/ Driverless Cars" ( ) JK--Designated drivers for college-town party animals?

RT @ClaireVinent: Predictions for the Internet of Things in 2014 & beyond  via @jameskobielus

"Ask yr data scientists 2 call my data scientists" ( ) JK--Re analytic products, they make/influence purchase decisions

Data monetization? Pay the persons for their personal data p1:  p2:  Tuesday #IBM quick-hit

RIP Pete Seeger. Played huge role in putting Americans in touch with their musical heritage. Profile in courage 4 resistance to McCarthyism

RT @dangillmor: Tom Perkins bravely wages his campaign to be seen as California's most clueless ultra-rich asshat.

RT @kexpplaylist #kexp Steppin' Out by Paul Revere & th Raiders frm Nuggets Vol 2 JK--Gr8 rockers undermined credibility with hokey costumes

"When to use Pig Latin vs Hive SQL?" ( ) JK--Good guidance for #bigdata analytics developers using #Hadoop #MapReduce

"Gluster Vs. Ceph: Open Source Storage Goes Head-To-Head" ( ) JK--Useful checkpoint on these communities.

#Criteria for high-quality, efficient #statisticalmodels begin with truth and beauty  by James Kobielus

A8 U don't need #data 2 prove every pain point but most of the time, ideas w/o #data are just conjecture. #CXO

A8: it's simply being ignorant. Not doing so allows your competitors a huge opportunity to poach your customers #cxo

A8: #cxo If you grow tone-deaf from not engaging in multichannel listening, you'll slowly wither. May not die. But may not grow.

A8: #cxo Where cust exper is concerned, U wont perish if dont have #bigdata. But you'll grow tone-deaf. Multichannel listening is key.

A8: #cxo Neither true nor false. Data scale not the issue. Competitive survival is all about foresight, agility, mgmt, & resources.

A7: Supply chain data on second- and third-tier suppliers can provide competitive advantage. #cxo

A7: #cxo Your data & your data scientists may B unexceptional. Your advantage may driving data-centric decision suppt throughout your biz

A7: #cxo Your data alone may not be your competitive advantage. Rather, your key asset may be data-scientist smarts for analyzing it all

A7: #cxo Compet adv is in unique value-prop difficult/costly for others to match. Your operational data may be unique asset. Mine it!

A6: #cxo In terms o churn risk, use data (survey, call ctr, etc.) to identify which high-CLV customers most likely to jump. Target them

A6: #cxo Customer survey and social influence data can reveal who drives sentiment, hence retention/sales/etc.

A6: #cxo Customer data can readily reveal who has highest CLV historically. Predictive analytics can help forecast CLV.

A6: #cxo Focus on a) customers w/high lifetime value, b) customers with greatest influence over other customers' behavior/opinions.

A5: #cxo Using data to pre-empt churn: a) analyze competitive price-points, b) assess own vs rival solutions, c) ID why U piss 'em off

A5: #cxo Pre-empting churn: a) reduce price, b) boost value, c) provide superior customer svc/engagement.

A5: #cxo Main churn drivers include a) you're too expensive, b) don't offer competitive value, & c) have alienated customer

A4: #cxo Use data you acquire during sales process to adjust your approach to woo each individual customer. 1:1 personalization.

A4: #cxo Direct marketing = targeting your message to the most promising new segments. That demands market research data

A4: #cxo Sales prospecting depends on data on qualified leads. That's fundamental to customer acquisition. Hot prospects?

A3: #cxo Perhaps leverage customer churn data. What "matters most" may be what the competition's selling.

A3: #cxo You should start by asking what overall experience "matters most." Data to answer that? Are prod/svc features priority?

A3: #cxo Of course, "matter most" starts w/ what they're buying. Purchasing data. Also, data on what prods/feats they're requesting

A2: #cxo Do real-world experiments & A/B testing of diverse variables across different channels. Measure differential #CXO lift.

RT @TabithaDunn #CXO A2: Ethnographic research into customer moments of truth can help prioritize what data matters most. #cxo

A2: #cxo Incorporate data on social influencers. Perhaps brand value perception shaped more by them than by your channels.

A2: #cxo Gather & compare data on customer interactions on various channels. Which channels drive sales? Which drive satisfaction?

A1: #cxo Data may be lacking to reveal cust "reachability" pref if you've never given them altenative. They grin&bear your current chan

A1: #cxo Data on past interactions can reveal if responding as U wish. But if have new custs or channs, past data not much help

A1: #cxo Data can help you determine which were the most EFFECTIVE ways to reach customers, in terms of response, acceptance, satisfaction.

@IBMbigdata #cxo A pleasure once again to be driving your tweetchat experience for the next hour or so.

Drafted latest #IBM jk blog: "Autonomic planet: distributed intelligence for a self-healing ecosystem"

RT @kexpplaylist #kexp So Says I by The Shins from Chutes Too Narrow JK--2003. Love this song. Have never understood the Sir Thomas More ref

Workday stress relief involves clearing my workspace of tasks. Do it, lose it. Like training with superheavy weights. Burdens test the heart

"ROI of Data Governance" ( ) JK--Good quantifiable metrics. What's a "tranche"? Don't get all French on us suddenly.

"Why video is next big thing in #bigdata" ( ) JK--Some #IBM brainiac called this "Big Media" ( )

"#BigData: Retailers, Supermkts, Med Mkts All Dive In 2 Extract Info From & About Consumers" ( ) JK--US lagging Europe?

Columnist (Net Wd, mostly), analyst (various), evangelist (IBM): that's the progression of my career as an IT industry opinion-slinger. Hmm.

"Social Media Data, Sensor Data, Open Data can make..." ( ) JK--Discusses #IBM & other #SmarterCities initiatives

"Using #BigData to Ask Big Questions" ( ) JK--Explore larger issues whose data-centric answers not evident 2 naked eye

"Wht Hse study #bigdata. 5 things it should know" ( ) JK--"3. Big means nothing." Bingo! Privacy vulnerable @ any scale

"White House Launches #BigData, Privacy Review" ( ) JK--Needs to engage UN, other nations in the discussions.

Morning's overstuffed inbox. Are newly received emails the crumbs U sweep off yr desk 2 start your day? Or are they the meat of what awaits?

Mornings are just matter of waking & slapping yourself back to the land of the living. Quotidian rebirthing. Hopefully absent the screaming

Morning is gray and freezing.'s January. I'll keep my expectations in check. It's a perfectly seasonable day, then.

RT @kexpplaylist #kexp Lost Boys & Lost Girls Club by Dum Dum Girls from Too True JK--2014. Excellent new song. They do Letterman this week

RT @kexpplaylist #kexp Bulldozer Love by Blank Realm from Grassed In JK--Cool. Sounds sorta like Tom Verlaine fronting Dandy Warhols

Smarter planet? Intelligence for a self-healing landscape p1:  p2:  Monday #IBM quick-hit

RT @kexpplaylist #kexp Salvador Sanchez by Sun Kil Moon from Ghosts of th Great Highway JK--2003. Brilliant grinding drone. Achieves majesty

Saw 2001 "Doc Martin" film that preceded TV show. Essentially, diff character, slightly diff name, same actor. Nicer. Prefer the curmudgeon

Well, that's two French presidents in a row who've split with their significant other mid-term. Surprised it didn't happen to the Clintons.

Another Sunday yoga teaches comes and goes, due to conflicting life priorities. What? This isn't their whole life? I'm shocked.

Met a longtime #IBM-er who was born & raised in Endicott NY. Sorta like meeting a Hershey Foods employee from Hershey PA. True Blue!

When some Facebook newbie posts in ALLCAPS, I'm wondering if it's from a Wang wordprocessor circa 1978.

Some called Newton's gravity "spooky action at a distance." You can think of quantum mechanics as spooky action close up.

Danny Kaye. "CBS Sunday Morning" profiling this all-around superb entertainer. Even his final TV performance ("Cosby Show") was memorable.

The news that IBM is disposing of its x86 server business to Lenovo for about $2.3 billion comes as no shock. 

Excited for tomorrow! Emceeing at #IBMConnect and get to introduce legendary @geoffreyamoore and the Queen of #socbiz @Sandy_Carter

Fingering my stubble. Sunday's one of my shaving days. Can't start the week without mowing the facial lawn.

Planning stuff I want to accomplish this year, work-wise. Hmm. Unusual for me over Sunday breakfast. Normally I drift at this time.

RT @shawnrog: 5 Laws of #BigData Startups - Analyst rant on Big Data startup briefings & strategy  #bigdata #startup

RT @TheDailyShow: Group of world's wealthiest people gather in secluded mountain enclave 2 discuss income inequality. 

The only Grammy contest that I have an opinion on this year is that Macklemore & Ryan Lewis should get Song Of The Year for "Same Love."

New jk #IBMDataMag article: " When and When Not to Have Faith in Statistical Models – Part 1" ( …)

WSJ review of WSBurroughs bio. I saw him onstage in Madison in 1984. Weird creepy unhinged incoherent ranting maniac. And that's being kind.

Can we drop the "also known as Lou Gehrig's disease"? This ain't 1941. By this time, every1 knows it's called amyotrophic lateral sclerosis.

"Stephen Hawking Proposes Radical New Theory of Black Holes" ( ) JK--Link is abstract of his paper awaiting peer review

My primary objective in consolidating many contributors' inputs to a document is to guard against broth spoilage. I keep tasting the soup.