Sunday, January 31, 2010
Week 3 Muddiest Points
I need to look harder at the formulas given in the week 3 readings and lecture, though Prof. He's breakdown of them made a little more sense. I realize they're important for this subject, and if I can learn College Algebra at 46, I can learn to at least interact with these, but it was a bit daunting the first time through.
Week 4 Readings
Reading 1 lists various aspects of a good searcher: good communication skills, logic and flexibility, self confidence, spelling, grammar and typing skills (for accuracy). These features of searching staff members (note searching staff, not users), with no-one really possessing all the listed talents, still aide the searcher in his/her tasks.
Reading 2, an early article (1982) describes 3 existing methods for searching: building blocks, successive fractions and citation pearl growing, gives their pros and cons, then adds a fourth: "most specific or least posted facet first". They state that this "interactive scanning" method, while taking longer, will yield more results. They also give advice for implementing any of these four methods, such as flexibility, searching more specific aspects first, and using thesauri.
Reading 3 wouldn't open on my computer. I'll talk with the help desk about that tomorrow.
Reading 4 discusses search terms and their effectiveness, including: written statements taken from users' comments, users' domain knowledge, relevance feedback from retrieved items, database thesaurus, and terms derived from intermediaries. That said, they discovered that there was little interaction with algorithmic approaches. Broken down into models on the user side (surface, cognitive, situational, and affective), and on the system side (surface, engineering, processing and content) brought better results. One conclusion is that interaction with IR is still a human art, and can be improved and enhanced on the computer end, but it will require learning from that human art.
This fourth reading I would like to spend more time with, to fully understand it.
Reading 2, an early article (1982) describes 3 existing methods for searching: building blocks, successive fractions and citation pearl growing, gives their pros and cons, then adds a fourth: "most specific or least posted facet first". They state that this "interactive scanning" method, while taking longer, will yield more results. They also give advice for implementing any of these four methods, such as flexibility, searching more specific aspects first, and using thesauri.
Reading 3 wouldn't open on my computer. I'll talk with the help desk about that tomorrow.
Reading 4 discusses search terms and their effectiveness, including: written statements taken from users' comments, users' domain knowledge, relevance feedback from retrieved items, database thesaurus, and terms derived from intermediaries. That said, they discovered that there was little interaction with algorithmic approaches. Broken down into models on the user side (surface, cognitive, situational, and affective), and on the system side (surface, engineering, processing and content) brought better results. One conclusion is that interaction with IR is still a human art, and can be improved and enhanced on the computer end, but it will require learning from that human art.
This fourth reading I would like to spend more time with, to fully understand it.
Subscribe to:
Posts (Atom)
