Skip to main content

Home/ VirgoLab/ Contents contributed and discussions participated by Roger Chen

Contents contributed and discussions participated by Roger Chen

Simple Middle Filter: All | Bookmarks | Topics
Roger Chen

Analysis: data mining doesn't work for spotting terrorists - 0 views

arstechnica.com/...k-for-spotting-terrorists.html

data mining reading research

shared by Roger Chen on 11 Oct 08 - Cached
  • Automated identification of terrorists through data mining (or any other known methodology) is neither feasible as an objective nor desirable as a goal of technology development efforts.
  • criminal prosecutors and judges are concerned with determining the guilt or innocence of a suspect in the wake of an already-committed crime; counter-terror officials are concerned with preventing crimes from occurring by identifying suspects before they've done anything wrong.
  • The problem: preventing a crime by someone with no criminal record
  • ...3 more annotations...
  • In fact, most terrorists have no criminal record of any kind that could bring them to the attention of authorities or work against them in court.
  • As the NRC report points out, not only is the training data lacking, but the input data that you'd actually be mining has been purposely corrupted by the terrorists themselves.
  • So this application of data mining bumps up against the classic GIGO (garbage in, garbage out) problem in computing, with the terrorists deliberately feeding the system garbage.
Roger Chen

The 14th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining - 0 views

videolectures.net/kdd08_las_vegas

conference data mining reference

shared by Roger Chen on 11 Oct 08 - Cached
  • Roger Chen on 11 Oct 08
     
    Videos of KDD'08 presentations
Roger Chen

Incredibly Dull: The KM Core Sample - 0 views

incrediblydull.blogspot.com/...km-core-sample.html

reading thinking

shared by Roger Chen on 30 Sep 08 - Cached
  • The Core Sample is -- like its name sake -- a snapshot of a point in time. It captures the various levels of "knowledge" and where they reside. The diagram also illustrates the rationalization and codification of knowledge as it rises through the layers.
  • Starting at the bottom, at the very core, are people. This is where true knowledge exists.
  • The next layer up is where that personal communication is expanded to allow people to "talk" to others they do not know or cannot meet in person.
  • ...3 more annotations...
  • The next layer up represents "knowledge capture". Here the knowledge is instantiated in documents of some kind: sample documents, lesson learned, case studies, white papers
  • Finally, in the top layer the captured knowledge and learnings are further refined into a defined set of templates, guidelines, and standard processes.
  • Collaboration strategies focus on the tacit knowledge layer. Methods like knowledge harvesting, lessons learned, and storytelling focus on the best practices layer. While ITIL, Six Sigma, ISO 9001, and other standardization methodologies focus on establishing institutionalized knowledge.
Roger Chen

用书的时候_文道非常道－梁文道的BLOG_新浪博客 - 0 views

blog.sina.com.cn/...blog_4c3782760100anu3.html

reading thinking

shared by Roger Chen on 29 Sep 08 - Cached
  • 弄不好，所謂的「多角度思考」其實就是一堆觀點和資料的羅列，而那些觀點和資料，不消說，全部來自互聯網
  • 一篇東西有不同甚至彼此矛盾的論點不一定就能顯示你懂得「多角度思考」，假如沒有一套邏輯清晰的架構安放它們的話，這只能叫做混亂，或者「短路」
  • 每個人處理資訊的方式都帶著點個性
Roger Chen

Data Randomization - 0 views

research.microsoft.com/...view.aspx

data mining papers reference research

shared by Roger Chen on 04 Sep 08 - Cached
  • Roger Chen on 04 Sep 08
     
    Attacks that exploit memory errors are still a serious problem. We present data randomization, a new technique that provides probabilistic protection against these attacks by xoring data with random masks. Data randomization uses static analysis to partition instruction operands into equivalence classes: it places two operands in the same class if they may refer to the same object in an execution that does not violate memory safety. Then it assigns a random mask to each class and it generates code instrumented to xor data read from or written to memory with the mask of the memory operand's class. Therefore, attacks that violate the results of the static analysis have unpredictable results. We implemented a data randomization prototype that compiles programs without modifications and can preventmany attacks with low overhead. Our prototype prevents all the attacks in our benchmarks while introducing an average runtime overhead of 11% (0%to 27%) and an average space overhead below 1%.
« First ‹ Previous 41 - 60 of 345 Next › Last »
Showing 20▼ items per page