Latest and Greatest in Assistive Technology

Session is in room Cairo/HongKong 10:00am

Presenters: Chan, West, Snider, Dias, Lay-Flurrie, Havlik.

Presentation1:

Intel reader http://www.intel.com/about/companyinfo/healthcare/products/reader/index.htm

Takes pic of printed page, converts to digital text, reads aloud. For sight disabilites. Camera on the bottom for usability!

Instructions are read aloud for the sight-impaired. It can spell out loud! People in user trials really valued this. Also, can change voice speed.

The accent is a little frustrating because of the disjointed words. 350 wpm is way too fast to understand. It sounds deeply amusing. But, says the presenter, power users can understand at this speed, and get frustrated if it's too slow...

The reader can create content for you. You can read a handout at a conference with it! Hold camera steady, take pic. It flashed! Camera flash ftw. Can take pic at any rotation and read it in. Algorith,s can take care of that for you- tell which way is up for the letters.

Capture station! Gotta turn the pages for it manually, but it does do two pages at a time. Reader has ~4Gb of flash memory. This is to improve bulk capture- i.e. for school, etc.

Key market areas: self-empowering students create their own audio books. Import/export files to/from other assistive tech devices.

reader.intel.com

Presentation2:

Voice reader. First, a video of a reproduction of a trial of the software. First issue: legal questions. Private business don't have to comply, but they lose lawsuits when they don't. (Did I understand that right?)

The tool runs on the cloud and is accessed via browser. It uses Dragon Naturally Speaking. There is a short delay in transcription because it goes phrase by phrase through the algorithm to transcribe. This makes silly owrd typos less likely?

This past Tues, the 21st Century Telecom law was passed. It requires captioning. It's a historical point for the American Disability Association. having a human transcriber is expensive, so they're hoping to market the automated transcription to customers of this law.

Presentation3:

Braile Writing Tutor:

for blind children in the developing world. COmes from Carnegie Mellon University called TechBridgeWorld. http://techbridgeworld.org/

87% of blind ppl live in developing world (because of health tech ungood)

It's hard to learn to read while blind.

Braile isn't a language like ASL; there's a different version for every language. To write, you have a tablet, put a sheet of paper on it, and emboss it with a little sharp pen/stick.

Slate and stylus costs <$10 in most places. It's the standard for the next 30yr, basically. So we have to work with it. They worked with a variety of schools for the blind. Bid problem: delayed feedback. Teacher has to read, then talk... and in at least one place, it's too humid in the summer to use paper!

They can't afford textbooks, so teachers just speak. If you don't know Braille, can't read your notes...

Automated Braille Writing Tutor startedw/ 2 grad students in robotics. This is v 2.5. Software is opensource on sourceforge! http://sourceforge.net/projects/tbwbrailletutor/

Audio feedback, buttons for little kids with iffy motor skills, curricular modes, game modes... "Power is the biggest issues in a lot of these places." Solar? Next version will have lots of solutions...

Example: Animal game. "Please write the name of the animal that makes this sound."

It's evolved to be the thing game.

There's a game that teaches you what buttons to press to write the numbers.

There's a strong accent, so the customize the sound file everywhere they go. very important for kids to trust the voice, so they usually use the teacher's voice...

Work with teachers to create different games.

There is a Kiswahili mode!! It sounds very cool.

I took some pics of the thing itself! Posted later.

Presentation4:

"I'm a deaf girl demoiing stuff for people with visual impairments."

Day job at microsoft is to lead customer service support sales division. Night job is this!

She's a very good verbal speaker for being deaf! She says that Microsoft tries very hard to get diversity.

"I am no techie. I have a music degree, people."

Windows 7 accessibility!

He sat down w/ disability ppl after xp and vista. Rough crowd, disapproving. Windows 7 had a solid goal to fix that!!! "improved hard."

How quick and easy it is to check your Word doc or powerpoint before sending it out. We don't measure disabilites... it's self-identified!

"Did you hear that? I didn't. I'm a deaf girl and I apologize. I'm also British."

WIndowsbuttun-U called usability features. She shows off the zoom mode! Windows is making standards for zoom and captioning... learn the shortcuts!

She made a slide that's difficult to make accessable. Accessability helper tells her exactly how to add alternate text!

"I'm being told that you can;t hear me." (crowd giggles) "I like how captioning will put my interpreter out of work."

Links will be on the Wiki.

There is a video! Promo and reassurance about windows 7 usability. http://sourceforge.net/projects/tbwbrailletutor/

Questions

Q: for IBM. Verbal transcriptor.

A: Yes. Watch http://www-03.ibm.com/able/ because it'll be in the "yellow zone" i.e. come play with it.

Q: re Braille. have you seen movie [what is this name] with the machine that let him read braille mechanically?

A: I haven't seen it but I know the machine. Refreshable Braille machines are difficult to make cheaply because of the large # of moving parts. People are trying to make them better now, but no win yet. In the dev world, people now type Braille!

A: from IBM speech transcript- debate re born blind children. If they do braille, they can master higher learning! Physics, chemistry. There's concern that current devices can't... thing... (huh? I got lost in the answer.)

Q: re localization to Intel. Visually impaired colleagues that use muliple languages.

A; reader is in English and British (accent) lol. Multiple languages are "on the roadmap" it seems simple enough but it'll take time.

Q: cost of reader?

A; $1500 for reader. Capture station $400. Tried to price it as cheaper than current solutions, whcih are ~$2000 (and not mobile!) Cost is driven by volume and they hope that they'll sell enough to drive the price down.

Q: ...

oops out of time

---
Note- who linked to this? There was a ridiculous spike in pageviews. I hope it was useful to you, whoever you are. :) Oct26,2010

notes on HCI track: 3 presentations. 3:00pm Thurs

Visualization data analysis of emails

mea18 from sfu.ca

Enron dataset. y axis = email addresses. x axis = time. Size of circles = size of email? Or perhaps number of emails?

Crowded areas in email set for six Enron people. Very clear high-traffic, low-traffic bands...

They wrote a tool to generate the graph from teh data (and assign/adjust parameters for view)

Centrality of person when it ocmes to emails sent, received...

Histogram of directly sent emails- peaks for one guy in Oct 01. Interesting!

Identify temportal gaps (events can be corellated to these). Can identify (from several addresses of one person) when they switched, etc.

One guy, four addresses. Can see switch patterns. Analyze size of sent emails.

Q: Advantages of visual = ?

A: See patterns in organization- easier this way.

(I think it's easier to see several data sets at once.)

Q: In the example w/ 4 email addresses for one person, was that autodetected? Or manual?

A: They're all parts of his name. Done manually.

Talk 2:

#hci

Judith Klavans and Jennifer Golbeck.

How CS can help museums with their collections.

This research is funded by the Inst. for museum and library science.

18 museums are users and test subjects.

Text written about images. Social tags form crowdsourcing.

User text is... misspelled, inaccurate, which language is it in...

"I call that an adjective noun, but parsers really struggle with this stuff."

Compare multi-word tags to Wikipedia titles to try to identify them.

Overlap of tags and terms (from professional to crowdsource) is 9.4% ...very low!!!!

"People use tags differently based on image type." - religious, ceramics, abstracts...

User looking for "red" is more likely looking for something abstract.

Research shows NO correlation between how often a tag is applied, and how high the image should rank in search results. Therefore, this implies that there should be other factors included.

Only 1 person in the room had been here at the first GHC! Yay growth!

steve.museum.org

Q: Maybe user training would give you better results? i.e. is your research disappointing?

A: Most ppl writing text about images are professionals. We're doing really well.

Q: I worked on this too, years ago! Our problem was that many objects were insufficiently described + messy description text (style and/or handwriting). Did that happen to you?

A: Yes. Database at typical museum does not allow additional data. Museums have different policies about data collection. (Why would a museum not allow information to be recorded?)

Q/A: Yes, interdisciplinary team.

"This is the profession of the reference librarians!"

"Semantic ontology, whatever you want to call it- we know it when we see it!"

Talk 3:

Icon design, consisting of rings with a color for price, distance, reputation, price(including shipping). size of ring correlates to higher quality. line new/used(black) at 0, 90, etc degrees. Map icons for competing items onto x-y. x axis = item's age (time since it's been posted) and y is relevance to the query.

Dropdown lists- sorting parameters, + box for showing details when mouseover.

Ran user tests: trained users, got them to evaluate the best item from data set according to given scenarios. There's a chart of right and wrong answers by how many- text, and icon. Some types of task did worse.

Which mode did you prefer? Mostly icon. Icon is good for quick overview.

Icon weaknesses: overlap = frustrating. color-meaning relationship wasn't intuitive.

Reorder rings? Make outer bits most important.

Q: Why rings? Confusing.

A: Prevent overlaps from obscuring all data.

Q: I think it'd be more intuitive to make them partial rings instead of changing thicknesses. What's the new idea that you mentioned briefly?

A: (Misconmmunication occurs)

Q: How did you decide which attributes of products to display?

A: Got 'em from eBay's sorting parameters.

(Black line indicates bidding rate of item.)

Q: Maybe less parameters?

A: We got the parameters from eBay. Smaller would be easier to understand visually, yes.

Comment: maybe different colors?

Exploring new Careers: US Defense and Cyber Security

Introductions. Everyone has very impressive titles.

Q: (very long, but upshot is "How did you get to where you are now?")

Sue (leftmost)- onsite campus interview w/ Lockheed. "I stayed in space my whole career." (And a whole lot more!!!!)

Rear Admiral Gretchen: My entire career is serendipity. ROTC 4yr paid scholarship. In return, serve for 4 yr. Submarine warfare, intelligence. Navy offered to send to school for Masters. Instead, got a degreen in (space something). Then, about to be CEO of comm station. Mentor = Rear Admiral Hite(sp?) and she said, go take one of the combat systems jobs. "Sixteen years in the Navy, never been to sea. I don't even know which end of the ship I should salute when I go aboard... You've got to be kidding" (not exact quote)

Reported to my first ship as a commander, and it turned out to be the most wonderful job opp I ever had.

(Comment by Linda: every time she says "managed the weapons systems" or something similar, I goggle inside. Like, the face of "zomg")

Information Dominance Corps, information warfare officers. (zomg!!)

2 women assigned to USS Grace Hopper (spontaneous applause)

Kathleen Hutson: 3rd from left: I grew up in a family of entrepreneurs. No boundaries, no ceilings. "It seemed like proposals were always due between Thanksgiving and New Years." Got a job with EPA. Worked. Met a lawyer from NSA, started working there! Got cleared. (clearance is needful.) had good mentors. NSA != what you see on tv.

Tamara Yu, 3rd from right: grad 6 yr ago in CS. Stumbled in genomics during masters degree. Worked on genome browsers. Went to MIT Lincoln Lab. Gotta ask questions, stay curious, go to seminars... "Develop both the soft skill sand the hard skills."

Lori Hecker, 2nd from right: oncampus interview "we can't tell you that" was the answer to everything. TI hired her in Dallas and Raytheon took over that. Where are these systems deployed? Always, unusual requirements. "I did something today that's probably going to save a life. Thank kind of job satisfaction, you can't get anywhere [else]."

"The mentoring is a huge aspect of success in this career."

Bonnie Kean. Rightmost, graduated 2008. Chem major, went to grad school in nanotech expecting to be a prof. Got to the end, realized donwanna be a prof. "I don't want to sit in a room with a hundred people who do the exact same thing as me and discuss it." Wanted to do something bigger.

Northrop-Grumman rotation program- different locations...

"If you want to talk complex systems, talk defense. They have to work 100% every single time."

They all seem to say- don't worry, and just take the good opportunities.

Moderator: parents == physics professors. Got masters in CS(?) and ended up in biz dev because of rewriting people's proposals. Won a couple of them. Mentor said "maybe you should do this for a living." resisted for a long time- wanted to be a scientist!

Open to questions from the floor:

Q: What impact has our career had on your family?

A: Lori- part-time for 10yr (since daughter born). -sometimes have to work more than the part-time week. Raytheon.

A: Bonnie- I move a lot. My husband telecommuted to Raytheon. Someone has to be in the supporting role...

A: Kathleen- husband == entrepreneur. Mutual flexibility.

A: Admiral Gretchen- I'm married ot a Navy Seal captain. No children. But you can! They just didn't.

A: Sue. Boss was curmudgeon. "Very unpleasant on a daily basis." Always approved them, and people usually came back early. Lockheed-Martin.

A: (moderator) Even when you think that children won't be needy... oh, my son is calling me. (He really was.) Group laughs.

Q: Mentoring... details?

A:

I was once reverse-mentoring a VP. It was awkward. I wouldn't recommend it.

Try to not be the mentoree of your immediate supervisor...

You don't need to choose the highest person on the totem pole! might not even be a good idea- big disconnect.

"Technical Women Network" (MIT?)

Lockheed: The tech fellows all have to have mentees. Online matching service for them, lol. Also, do peertopeer mentoring. You can really help each other out.

Blue: can be a many-to-one relationship, too.

Q: if you're overwhelmed by options, what do you do?

Admiral: Look inward. What really motivates you?

"trying to get a piece of gear fixed..."

Moderator: does your uni have a mentoring program where you can look up willing alums? (Linda comment: hahahah, not my university. Doe we have willing alumni?) She gets 5-6 phone calls/yr, + more formal mentoring program.

Make one. Then if it doesn't work, give proper notice do your best until the very end, and get something else. Now you'll know something new...

Q: (voice is almost gone!) How do you give back- advocate for others? Mentor for others?

A:

Well, the mandatory methroing programs relieve the guilt about not doing it. Easy avenue!

"enlisted in the organization" zomg.

I was at a conference... one was a reporter who followed some of the prominent women in the news. She told us about what she found important in their leadership styles. It was: providing opportunities for others to network. Connecting people with each other.

Set up your own environment to network and share.

end

"Exploring New Careers" HanoverAB 2:15

Exploring New Careers: Choices in Finance, Online Information Services, and Small Business Technology

Hanover AB 2:15
Women speak: Alternate tech career paths.

Companies represented:
Intuit
"DoYch-bank" (spelling?) Melody
Goldman-Sachs (originally from Dublin)
Hannah Darry from Black Rock, based in NY, manage about 3trillion total. ~20% of firm is in tech. Other banks use their internal-use software. "I manage software teams in 20 countries around the world." (did I really just hear that? Zomg.)
Christie Claude from Thompson-Reuters. 50,000 employees. 15,000 of those employees are "technologists" (Yayyy, new buzzword. :p) been there from 22 yr, started as a programmer ("developer") and now in charge of stuff. "I have ~engineers in my department, and we develop the products that the legal department uses."

About ~1/2 the people here are students. Zomg.

Question to panelists: How does your company make the world a better place?
Intuit: saving time for small businesses, increasing tax returns (talking about Quickbooks), making more profit margin. Also: 1st carbon calculator footprint for small businesses.

Goldman: program "10,000 women" 5-yr investment by the firm to teach biz management and entrepreneurial skills to women who don't otherwise have access to all that. Also "Community Bridge Program" to 2-yr universities to get them to get a 4-yr degree in CS - they get a 6wk internship in Goldman. Also, "Community Team Works" program: "engage with local orgs to do some sponsorship activities..." people at each office take a day off to...

Thompson-Reuters: likes this question. "This is why I've been in the online information industry for 22 years." "We make a product for law enforcement officials and government agencies." Can know when a person has lived at the same address with another person for a month- so there might be a connection... (this is creeeeepy from a personal security point of view) Used this product to track down and rescue a kidnapped girl. Cool!

Question: (I think it's "What keeps you in industry?" or perhaps "How did you get into industry?")

Black Rock: The fact that what I do applies to real life. Also, the technology is interesting. Retirement accounts, etc.

Doychbank: Haven't written a single line of code in 30 yrs in the industry. Left uni in a terrible downturn. Commerce degree, accounting minor. (From Australia!) Got her CPA, eventually decided didn't wanna. Got a job with Leiman Bros. Worked on tech. implementations. Went to Sydney, got job at current company. Bigger and better. Moved to NY, lots of travel.

Q: How is the indutry changing over time and how does it affect future jobs?

Intuit: Not only desktop software now. 1995, 2% of Intuit customers were internetish. "My biggest challenge was getting people connected to the internet." Data mining is a huge challenge. We have lots of data. We can get wonderful insights. There are tough technical challenges. 18 mobile applications.

Black Rock: Securities lending. Building algorithms for automated trading. making sure that there's no funny business going on.

Thompson-Reuters: Legal industry- two years ago, our lawyer customers would charge by the hour. They didn't want us to make it faster and better. Those days are gone. I think that lawyers are still adjusting to that. It's fixed bids on tasks. SO we need ot make products that can help lawyuers find what they need quicker.

Q: What cool, cutting-edge tech is happening at your place?

Doy: Building desktop app for global markets- trading sales priducts debt equities (this is a foreign language to me) "This is a great app; it's a pity he won't use it, because he never sits at his desk. His assistant prints stuff out and carries it around with him all day." So I told [some important guy] about this, and we had been talking about the Kindle (which hadn't come out yet). iPad wasn't even announced yet. But they built something based on the rumor of the iPad, using the iPhone devkit, and assume that it will be portable to the iPad. As soon as they released the OS, we ported it. Bought the first ones. In the first week after iPad release, we had a prototype and the guy could have connectivity to all his data. "It was pretty exciting to be so far ahead of the curve."

Goldman: "My whole world is about speed, performance, low latency." 100s of thousands of transactions every day. I recruit a lot of people form very tech backgrounds with no financial interest. but they do what needs to get done! Also, the fin industry "works back with" the different software firms because they consume and use so much tech.

Q: Security- what risks have been introduced?

Intuit: Gotta be up all the time, even under max load on Apr 15. LOTS of people wait until then. Cloud computing is a security risk as well as a scaling solution. "Unstructured time" to help solve unknown problems. There is a thing called snaptax- it's for the iPhone- you can take a pic of your W2. http://techcrunch.com/2010/01/04/snaptax-iphone-turbotax/

She mentioned Ruby on Rails!!!!!!!!!!

I'm notcing that she's trying to avoid specifics about "security risks" ...

Thompson-Reuters: We screen all of our customers. We only sell to law enforcement, CIAs (plural??). They don't sell to stalkers or most civilians. They monitor the usage logs- if there appears to be a batch going in, then we investigate and shut that off.

Q: Mistakes you've made

Doy: Doing a migration from one trading system to another. Practiced a million times. it was loading slowly, so hit the button again. Which ended up loading all the trades twice. It's the middle of the night. All of these ops people in London are waiting for me. What shall I do? Had to cancel a couple thousand trades that night. Worked through the night. Afterwards, they had respect for her (and it was fixed by morning!) Didn't set back her career.
Let the horse out the barn? Go get the horse, put it back. Be a problem-solver. Take responsibility. And, well... no one's going to die.

The other thing: spent a lot of time in the career trying to fix things that I wasn't very good at. Now I'm like "Why did I do that? Should've just worked out what my strengths are and made that my skillset."

We had an astronaut come into the office last week- he helped fix the Hubble. At the end of his speech, he said "just do what you're good at."

Moderator: Women point the finger towards ourselves more often than towards others. "We fall on teh sword sometimes excessively. ...Be honest but don't be self-flagellating."

Intel: Should have taken more opportunities earlier in the career. Learn more (about yourself.)
(Linda's thought: do men worry about learning about themselves this much?? I think not... or at least not out loud. How come?)

Thompson-Reuters: I wish that, earlier in my career, I would have asked for more things.

Goldman-Sachs: Encouraging women to stay technical for longer in their careers. Don't take your fingers off the keyboard. Women too often go into management.

Questions from the audience:

Q: What mentors did you get? How?

A:

Doy This guy- was new in from other company- told it like it is a lot.

Intel: Ask for help, even if it's hard. Go into the situations. Ask for feedback- how did I show in that meeting? How did the presentation go? Be open and ask.

Q: Are you programming for Android? Where do you think that that market is going?

A:

All new apps. Mobile first (Intel) "There's a lot more cellphones in use than desktops." including text-based.

Q from Moderator again: can you map what you learned in school onto business?

"Did we learn in school? I thought we just drank beer" (from onstage! heheh. fun times.)

A: "Well, I can talk about how it was different..." Impact on more people. "Well, you only get half your paycheck." Workforce: so much work, no semester break, all self-scheduled. No "class twice a day" anymore. "I've learned how to learn." At Intuit we have tech forums, learning development stuff... "It's going to be different in three months anyway." Eeeeevery day.

Audience Q: My background in in math. Only a little CS.

A:

Thompson-Reuters may hire you.

Also: (from Doy) How do you get the gold out of the mine?

Oop, call time.

PHD4: networks/machine translation (notes)

Heuristic word alignment with parallel phrases
Maria Holmqvist
Sweden. PhD student

Word alignment applications: bilingual dictionaries, computer aided translation, statistical machine translation (most likely answer)

Corpus (training sample?) - hopefully millions of sentences w/ translations.
Extract phrase translations from word-aligned texts
Estimate translation probabilities (of different interpretations)
This is unsupervised- does not require human intervention for training.

Gold standard evaluation: % correct links in sample.

Alignment with parallel phrases (Maria's method)
- uses manual alignments to improve alignment quality.
-- find common alignment patterns, match them to patterns in new text.
- generalize phrases with parts of speech
-- replace nouns with noun tag, verb with verb tag- i.e. creating grammar translators!

Limit number of part of speechtags.

Questions to answer with the research:

How does phrase-based compare with statistical alignment?

How can they be combined for good results?

Ongoing work:

comparison of the two methods

comparison of wars to combine different alignments

machine translation experiments

Q: What kind of work has Bing done on getting diverse data? Your data has been academic. How do you verify your data?

A: Verify by comparing to Gold Standard of "manually aligned"

A: On the web, can get comparable text but not parallel text. So it's hard to get and work with "real" data. But work is being done to collect more diverse data.

Q: Have you looked at what the statistical method gets wrong?

A: Infrequent words. I'm doing a deeper analysis. Problem: in Swedish, it's a compound language, so there will be lots of previously unseen words in a given chunk of text.

Ewa Musial

University at Albany, SUNY

Modeling and testing of services-oriented applications

Existing testing methods: single-tier testing (presentation tier, application tier, service tier, data tier)

New approach: choreography/orchestration testing- "web flow in application layer"

- about "Business process execution applications"

- they don't question the incompleteness of current techniques.

Problem: lack of techniques to test ~~

Test model:

~

test study 1~

test study 2:

fault detection by users. 36 test cases, 31 faults.

results:

results show that approach is effective for improving reliability of industrial projects

develop complete tools for automated testing

Q: Just java?

A: yes, but not always

Q: (from guy!) Can you characterize the general faults that you were finding?

A: Yes. People tend to test easy things. If people move, the database won't keep up.

Q: How did you make the complicated dependency graph? Can't all that change?

A: It can change. Sometimes it needs to be adjusted. During testing, the reliances change...

Fariba Khan

U of Illinois- Urbana, Champaign

Assuring network service with bandwidth and integrity-based fairness

Also, DDOS attacks.

Fair-queue- head of line blocking

IP spoofing- can't use IPs as a solution

Graph- testing packet success rate under different attack circumstances. Attack, no defense and Attack, FQ, spoofing were both ~4/100 packets getting through.

Ingress filtering is neither complete not verifiable. IP of a filtered domain can be spoofed either in the same domain or in an unfiltered domain.

Design: integrity levels, integrity token, queue.

Spoofing index table (?)

MAC integrity token

(My worry: would this make communications more tracable? Where does this interact with trueanon and cyberprivacy?)

Q: How did you develop a relationship with an ISP so they will use your algorithm?

A: Haven't yet.

Now: turn in your presentation evaluations!

New Investigators 2 - Data Management (messy notes)

Zhara Ferdowski
Data mining to find socioeconomic character of modern cities

Location: Dunwoody 9:30

Chicago modern data. (survey from 2000)
K-means clustering

data mining vs. statistical techniques.

building permits- studying renovation leading to gentrification (is that the right cause-effect?)
also using the 1980, 90 census data.
Data preprocessing- all sets in different formats.

Chicago - 860 census tracks (what is a census track?)
Zhara sorted them down to five clusters.
1. mainly hispanic. renters, low income, low edu.
2. mainly white. middle class.
3. high income.
color map- slides. orange (3) = north side.
4.
5. blue (I guess- poor?)

Questions:
How did you determine the # of clusters?
A: tried # of clusters from 2-20, based on 2 criteria. got opinions of experts on which is more "real"

Q: What was the hardest part of the proj?
A: collecting data. there was no consistent data... different formats... not complete data set.

Divya- PhD @ Penn.
9:48
advisor: Trent Jaeger(sp?)
Retrofitting Authorization in legacy systems

full room!
X server system
mediation statements-
- declassifyers(don't put in plaintext- encrypt before writing to file (entering password))
- type 2??
- access hooks: user must have permission to read/write to file inorder to use...

example: logrotate (code! also small red lines.)
goal: automate the process to re-secure old programs or at least detect fails.
security lattice- secrecy, integrety, authoization.

information flow graph. (model program using one.)
place mediators at min-cut vertices.
"completeness guarantee"

can mark lib files as "must not" locations
if there's already a bit that encrypts, all flows through that can be dismissed.
heuristically better mediators: avoid putting them in functions that are accessed often.

applied this to ~8 programs
profilecode at runtime to gather this info.
hoisting: reducing field sensitivity

trying to write it in C. Not done yet.
example: tinySQL

Q: the program wouldn't work in an environment where you're spring-injecting variables, would it?
A: static analysis tool. Can be used to retrofit exisiting programs. (Was this an answer?)

Query-based data mining for the web
Barbara Poblette (researcher at Yahoo. in Chile, santiago, barcelona, etc.)
info retrieval, data mining.

High data dimensionality. Focus: user behavior. usage logs. (how did she get this? eep.)
automatic organization is important for IR

"Right now they have human editors, and that's not a scalable approach."
2" query dociument model, ___
traditional: vector space model i.e. bag of words
query model: bag of uery terms
query-set model: bag of query sets (how of you determine a query set?)

query - transaction which happens n times.
also, # of clicks from query to doc.

2 types of web graphs: click and hyperlink.
link graph: susceptible to spam(?)
combine them, find best-of.
"Queries are an excellent way to reduce the feature space of the web."

Q: implications = ? how did you gather this data?
A: Interning at yahoo, access to yahoo access log. "they have all the queries..." big log... 144 mil documents...

Talk 3:
thesis
relational data mining.
classification of
"A new link-based naive bayes classifier"
Bahareh Bina
Oliver Schulte

scenario 1: social network. goal: predict whether a person smokes. check smoking habits of friends.
scenario 2: goal: will customer buy product x or no? tradiationally, use only attributes of customer. however, now can use attributes of their friends sigh.

link-based classification:
predict the class label give attributes of obj, related obj, and of the links
(marketing :p)

prev work: graph-NB
all assumptions together = standad relational NBC.
our algorithm: LNBC
consider weight and number of rows (to normalize the weight)
improve classification accuracy on benchmark datasets

Poken setup

So, when I was uploading pictures to my computer, I switched over to my Windows partition because F-spot was frustrating me by not being Picasa, which is frustrating me by not being installed yet.

So I plugged in my Poken in Windows and the folder had three files: start_poken, help, and autorun. I opened the start_poken.html and it took me to the login/create account page. Logged into my existing account. Got taken to the "dashboard" which showed the people I had pokened! (Pick your favorite pokenesque verb. I prefer the self-referential "do what pokens do.")

Uploaded my headshot (same one as on this blog- it's recent, looks like me, etc). Stared at the card. it looks very... bland. But it has my info. Close damn enough for tonight.

I'll probably do this all over again tomorrow in ubuntu. For now... fun times!

GHC: Tuesday! (job fair)

Pics here: http://picasaweb.google.com/wordymusic/GHC10Tues#

31 pics, creative commons. email me at wordymusic@gmail.com if you want something taken down or want something that you think was taken to be uploaded.

It's going to be even more crowded tomorrow. Today was just the job fair; tomorrow all the industry professionals will be here, and it will be epic. I remember this from last time.

My hotel room wifi is pretty atrocious: it's ranging from 40ms ping to 120 ms ping, and from .78Mb down to .25Mb down- which is to say, not as good as the internet that I was getting on the plane ride here. So I'm sad about that, but it's not like I have time to watch videos, so I guess it's okay. Also I think that bandwidth was much higher this afternoon because not so many people were on it... which tells you a lot about the network configuration, actually.

All three of my roommates are asleep right now, so I'll try to be brief. (And disjointed, because I took notes sporadically throughout the day.)

This evening around 8:30, I went to the Hard Rock Cafe with Mariya (roommate) and two people from her university, and we chatted and chatted... turns out that they're all three PhD students! (I just finished my undergrad degree.) So that was interesting. I drank some margarita that the bartender came up with- pomegranate and melon, I think? It was tasty but very mild.

This morning at 12:30pm (dude, it was totally morning), I reported to my Hopper shift! I signed in, picked up a red vest, met my fellow Hopper Juliana, and hung around the poster table until three pm, talking everyone's ears off. Juliana is awesome- goes to UW (giant university in Seattle- an order of magnitude more undergrads than IIT) as an aerospace engineering major (actually, a pseudonym of that) and her mom works for Microsoft and has been trying to get her to come for years...

I considered begging an introduction to her mother, but I'm not sure that I want to use .NET all that badly. .NET is a lovely lovely program, but it has a high learning curve, which I am currently on the wrong end on. And frankly, Ruby is more fun. Hell, Java, C, and C# are more fun. I wouldn't know about FORTRAN, never having used it.

Sorry Juliana, if I embarrassed you. Watch me do it to other people now too!

While I was being a Hopper, I met this lovely older lady who's one of the official photographers for GHC10! I showed her my Rebel xT and she told me that she was a Nikon lady... actually that her favorite cameras were this German line of hardware that's 3x more expensive than anything whatsoever... and she told us an excellent story about when she spoke to her granddaughter's highschool photography class and told the teacher that he(?) was full of shit for teaching darkroom techniques but not Photoshop. I love it when industry professionals tell off teachers.

There was a man who works at the Hyatt who, after he walked up to us (at the poster table) and conversed with us re: a pile of boxes to be taken away, the first thing that came out of my mouth after he was out of earshot was "He... has a very purposeful stride." It was very brisk. There are actors somewhere in Hollywood who wish that they could do that.

Moving on. Some nice lady came up to me, said "Are you Linda?" and proceeded to tell me that she was "geekettebits" and I said "Oh yeah, I friended you this morning." Man, I <3 geekettebits ="=">

"My attention span is about three clicks long." That's in my notes. Not sure where it came from.

As I was sitting at the table, Zara and Rachel came by wearing a massive amount of Google swag, and chatted for a bit about work at Google and all that. Apparently the guy in the wraparound black and white Google t-shirt is the marketing guy, is named Damien, and ran the meeting at which they discussed what kinds of swag to have...

I met Yang, who is a Hopper and is presenting two posters! As far as i could understand, one of them is on paralell computing with biotech (to, or with? I don't know) and the other is on how to do a server migration while also using parallel computing (is that a solved problem, or still under debate, or...?) ...Anyway, I plan to go check out the posters and take pics and generally be very impressed. I'm theoretically doing a poster next year... depends on co-authors and research outcomes.

By about 5:30pm, I had talked myself hoarse. So I got some mints and a bottle of water and kept talking. I was the girl in the purple shirt who talked your ear off. Yep, that was me. Follow me on twitter (wordymusic) so I can do it there too.

My roommate Mariya worked a summer at some cloud computing company called Scaleright or Rightsize or something- lots of details, very cool-sounding. Man, there are so many places that it would be cool to work at!

I was planning to do a detailed blog with pictures about the swag bag, but it's past midnight and I'm going to be up before 7am tomorrow, so instead I'm planning to do an end-of-conference swag post about everything that I pick up all week.

By the way, the object persistence framework Hibernate is rocking my reading list right now. It's so cool! I'm going to set up a mini implementation. ...But not tonight. For the love of little green Poken, not tonight.

I went and got my Poken! Had to pay $15 for it, since I'm not actually a student any more (barely!) and I chose the little green one with three eyes... so cute! I used it several times today, so it's totally worth what I paid. I haven't set up the online account yet, but I'll probably do that tonight. I hear that it might be a little complex...?

Next to the Poken table, there was a table with hardware on it. Which I looked at and went "ooooh!" because I used to assemble stuff like that, so I asked the guy sitting there what it did. It turns out that his name is Spencer, and he had two smartphones, some hard-wired switches, some radios, and two step-down power devices so radio 1 wouldn't short out radio 2 at such close range... and his company Rayetheon is making a system so that mobile phones are usable in war zones (unmodified cell phones are very locatable and therefore dangerous/useless in the field). We talked for a while and he gave me pieces of paper and we discussed Fort Worth and the programs there... I'm not sure that the commute would agree with me, but they're doing some rocking things over there. I like travel, but not between coffee and meetings. More like between where I am and somewhere interesting.

Ended up having dinner (meh, box dinner wraps) with TeeTee from IIT and some people that I just met... fun times!

I stopped by the opensource table but it was full up, so I headed out. I'll totally be back, though.

I should look at some of the grad school tables, since it's possible that I'll be gradschooling relatively soon.

On Friday from 3-4pm I'll be at a lecture by someone from ThoughtWorks - it looks interesting! Also increases my chances of being employed, sent to India for training, and being generally incredibly happy. (I've never been to India! Dangling a free international trip in front of me motivates me impressively.)

I ended up talking to some excellent people from the Xetron division of Northrop Grumman, which is the company that all of my engineer friends slaver to work at... and it sounds like it's pretty excellent over there, so I'll definitely talk to their website and forms and all that and see if anything comes of it. (Because how cool would it be if I could help some of my friends get jobs? Like, just by telling them that they should bloody apply, because I work there. That would be rocking.)

I talked to... some company whose logo looks like Picasa's, and whose acronym sounds like the "jitsu" part of ju-jitsu... anyway, not seriously considering working there, mostly because they sound like they're kinda stiff in general- but I was talking to one of their HR people about their fiberoptic cable network analysis hardware and I would really really like to talk to one of their technical people about how that all works because I'm a brat with an oversized curiosity. I'll poke their website and see what there is to be seen...

And I talked to other companies but frankly it's getting towards 1am and you don't care anyway, so au revoir.

I was so good at the job fair that I only have one resume left. Gotta stop by the Kinkos upstairs tomorrow- not that I really need so many more, but they're nice to have.

And the codeathon is Thurs afternoon! Come. I will be there, writing bugs. (Code is by definition a series of bugs, right?) http://eden.sahanafoundation.org/wiki/GHC2010

I have the feeling that my post-conference followup will take weeks.

Dammit, one of my roommates is snoring. I won't tell you which one.

First ZOMG GHC moment

So at IIT last week I was chatting with a nice lady from ThoughtWorks, and we talked about how we'd both be at GHC soon (and there was amazement over that!).

So I'm sitting in the lobby because I'm not checked into a room yet (long story) and she walks up to me with a coworker and says hi! And he have a happy enthusiastic conversation about the programming books we're all reading, and my network analysis on the plane wifi, and the t-shirt that she gave me that I modded. And she's going to send me "the survey" which is part of the potential-hire process. And I am extraordinarily happy because this makes THREE companies that have me on round two of their hiring process!

It still remains to be seen whether any of them will hire me. I'm actually financially decent for now, although I would really love a change in circumstances. And I fully expect to find amazing possibilities at the GHC fair. Getting paid to do what I consider fun? I could be so lucky!

GHC: where will I be when? Lastminute planning.

The RSS feed is http://geeklinda.blogspot.com/feeds/posts/default/-/ghc10 but I don't know how to add it to the list yet. Finding out now/shortly.

I'm in the airport in Chicago waiting for a plane to take me to Atlanta, where I'll stay a night in a cheap Best Western before going to the GHC hotel in the morning and meeting my roommates!

My current plan (yes, it's sparse; but I'll be going to a session every period- I just don't know which ones yet.)

Tues morningish check into the room and drop off my luggage (and leave a not for my flatmates, who're landing at 11:30, I think...)

Tues 12:30 work a Hopper shift at Shirts and Posters

Tues 3-5pm say hi to Bj at the community table in the Cyber Center.

The sessions that I'm assigned to cover:

Linda Goldstein

Blogging, Note

taking

1. (*) Susan Graham: Using Information

Technology for Health and Healthcare

– A Look at Research Challenges

[note taking and/or blogging]

2. Distributed Computing Systems

3. (*) Change Agent Awards Panel

I don't know what times those are at, yet. I'll look it up first thing on Tuesday. My laptop isn't dealing well with the formatting of the GHC master schedule, whee.

I'm not entirely clear on where I'm supposed to blog about these sessions that I'm covering, or what the difference is between a blog and note-taking. But I'm good at taking notes, and I sure do blog a lot, so I figure that I'll ask BJ or whoever's around about what format they want everything in.

After four years of university, no note-taking assignment scares me.

I'm really looking forward to having a Poken! I registered on the Poken wenbsite already but I don't see a way to reserve them on the GHC website. I'll be there early, so I guess I'll probably get one if they have any left.

My actual Hopper shifts are:

Shift #1: Tuesday, September 28

12:30pm-3:00pm

Shirts & Posters

Shift #2: Thursday, September 30

6:30am-12:00pm

Cyber Center

I might volunteer for more of them, if there's a time slot where there isn't a presentation that catches my fancy. Working the table seems like an excellent way to meet people.

I'm also excited because I bought some new clothes that are pretty spiffy (by my usual grunge college student tracing-cables-into-dusty-places standards) and I think that I look pretty excellent. I totally have tennis shoes on, though. New black ones, but still not heels. There's no reason to wear heels, you know... well, no reason that I like enough to use. Uncomfortable and a little dangerous: not how I like my walking experience.

I'm looking forward to the career fair and seeing the lady from ThoughtWorks again, and registering for the cool competitions that some of the tables are holding...

To be honest, I'm looking forward to fitting in. As a female geek, there aren't that many places that I can do that irl, not really. It's kinda what's been getting me through the week.

Right now I'm reading the book Coders at Work, edited by Seibel. And it is fantastic. http://www.amazon.com/Coders-Work-Reflections-Craft-Programming/dp/1430219483

Jobsearch

Yes I am looking for a job. Coding, preferrably, although I admit I'm a fucking noob. However, I'm a noob with decent design sense. Principles of refactoring are not foreign to me. UI design is fun. Database structures are fascinating.

My LinkedIn is here. http://www.linkedin.com/in/lindagoldstein

I'm basically counting on GHC10 http://www.gracehopper.org/2010 to get me a job at this point, although if anyone would like to save me the trouble and offer me something nice, that would rock. And it would mean that I could settle in and enjoy the GHC tech track sessions without worrying about going to "birds of a feather" sessions.

I understand that networking is important, but I can't bring myself to respect the principle. Networking is about meeting people, learning about them and who they are and what they control, and connecting them with each other (hopefully though you) or connecting with them yourself (somewhat awkward, eh?) in order to (mutually, if you're doing it right) benefit.

Anyway, right... I plan to blog about the sessions that I go to at GHC here, and link to them via twitter with the #ghc10 hashtag (I am such a social networking schmuck) and hopefully introduce people who like hiring people to the concept that I am somewhat intelligent, love to learn, love to do stuff...

I definitely plan to be running Ubuntu during GHC10, and therefore I should reinstall it (I uninstalled it last night because, (see previous post) it was a botched install, albeit one that helped rancorbeast evolve a pretty excellent solution to a prevoiusly unsolved bug.)

I would like to point out that my jobsearch is having some minor success at the moment, although I would much prefer major success. -shrug- I'm good for rent and groceries for this month without dipping into savings.

...Oh, and I also bought about ten books used on Amazon. Some scifi, some programming. Fun times.