December 06, 2007
12/6/07 Blog #8 Course Recap
What was BIT 330?
So the goal in taking this calss was to improve our search and trapping skills on the internet using a variety of different techniques. Some of these however resonated better than others. For instance, I learned a ton about RSS feeds, manipulation of them, etc. We also learned alot about page monitoring.. but this was just overkill. I've still never had a situation where I think I might use something like this. And if I do.. I will get back to you on this one. In other words, we went in depth and some of that depth wasn't such interesting or fun depth to be in..
So what tools were most interesting?
Feed 43 was actually one of the coolest tools that we learned about. Ironically, I still haven't put it to use because it's got a pretty steep learning and patience curve. However, I really enjoyed learning about its functionality. Above all, it really showed me that there are some very advanced trapping tools on the internet and that I wasn't looking or thinking hard enough to find them in order to take full advantage of their capabilities.
Google?
Yes, we learned about Google, but also Yahoo, Clustering Search engines, Blog Searches, RSS Searches. I think we all grew up thinking the quotation marks in the search was pretty cool. How about using some even more advanced syntax. That's right. Before taking this class, I thought that all of Google's searches were pretty much all pre-canned stuff. Well that's also true, but you can type up to 32 entries to Google to make a very very specific query. And the cool part is.. if that actually exists, Google will find it. It's not like it sees 32 pieces and thinks "shit how the hell do I do this.." then returns 0 results. I think this is absolutely fascinating.
What to take out?
Take out!? No. Haha. One suggestion is to take out ONE of the TWO term projects. Two is just overkill. I spent alot of time on the personal one (because it was first developed) and I really appreciate the understanding in allowing us to weight them differently. However, maintaining two is just overkill especially when we're using the same tools on different topics. Maybe separate tools for each of them?.. depending on the time in the semester where they're introduced.. make them smaller in scope? I think this may work in the future.
Posted by asmuench at 12:17 AM | Comments (0)
November 12, 2007
11/12/07 Blog #7 RSS Recap
I've been pretty impressed with the power of RSS feeds so far this semester. Granted, they don't always give me the information I'm looking for, but they're a pretty easy way to keep up with things (even when I'm being lazy). The reader that I've been using is Bloglines and it does a pretty good job of keeping track of my feeds, although I like to delete stuff that annoys me (feeds that return data I'm not interested in at ALL)
Which feeds have I been receiving
I'd like to do this a little differently and describe the types of feeds I'm receiving so far this semester. OK, before you get all critical of me you should know my personal topic is on Microsoft. Ok, now you're ready to hear about this: Google and Microsoft. Yup--those are the two things I've been taking feeds (and deleting feeds on). I started by searching the bloglines site for different feeds, and then later in doing my term project stumbled across certain sites that just weren't listed on the Bloglines site (i.e. different Microsoft.com feeds, one for press releases, one for top news). As for Google, I've also used RSS Micro to find some great blogging sites. In an earlier blog I wrote about the different feeds, you'll see other engines I've used--albeit without much success.
What are they
To be honest, most of them are news about the company. I guess you could call them blogs, since the official sites allow for posting of comments--but they're usually news stories. A few feeds I receive are blogs and they're straight up *weird* but very amusing to write (and NO that was NOT an attempted bold syntax in Textile--ohh how I amuse myself). Actually, one of the best feed that I've been receiving is a modified one I've done that combines many feeds into one using Yahoo! Pipes. It's a very easy to use interface that allows users like myself to manipulate feeds. Oh yeah, and if you haven't checked out my YouTube video, you can see it on display in an earlier blog entry.
Andrew, let's be honest, are they helpful?
Mmm.. Gotta love the questions that are best answered "depends". To give an answer for all of you who hate this non-space Yes, they have been helpful. The reason I was using RSS feeds (other than the fact that they were required), was that they kept me up-to-date without really having to do much work. This is exactly what they did--and so they lived up to their promises.
But since you're a computer science genius Andrew, what would make them better
Wow. I'm glad you asked, because I was just about to tell you. What I've noticed so far is that RSS feeds still are fairly difficult to filter if you don't use second-party software like Yahoo Pipes (mentioned above). Essentially, you can't take one feed and parse it quickly (Pipes isn't that hard, but there could be an easier method. A good example might be a query that the feed flies through (some sort of standard that's included), and then that filters the RSS feed.
Also, I wish there was more feedback on the feed itself as opposed to items ON the feed. What I mean by this is that usually people are allowed to comment on whatever the news, blog etc on the feed returns. However, (and as far as I know), there's no way to post a comment about the feed itself--rate it on a uniform basis or anything to this nature. I'd like to see this in the future, so instead of relying on algorithms to return me the best feeds when I'm searching, it can be based on user reviews instead.
Posted by asmuench at 02:50 AM | Comments (0)
November 05, 2007
11/5/07 Blog #6 Email Alerts
Mmm..
The overarching ideas here were to use the same "basic" queries to see which one returns better results than the next. However, I wouldn't recommend using "Microsoft" in your query, unless you've got a nice email filter setup. Other queries I've used are or are variations of "Vista" and "Xbox", in which case Google and Yahoo! dominate CNN in many respects. However, CNN can be an RSS feed, which you can port to Pipes! and customize the heck outta.. which I still haven't done, but would recommend.
From Where: Google Alerts
What Are They: OK, so my personal topic is Microsoft, and I decided it would be most valuable for me to put this as my choice for email alerts.
How Helpful Are They: I started off by putting "comprehensive" when I set up the alert in a hope to understand how it worked. However, as I soon found out, this returned wayyy too many results (I had it coming to my email box daily as well). So I changed this to just "News" and then I had filtered for results that first included the new "vista" operating system. I then started to think I could sort the news this way instead of doing it on "web" setting and specifying "news" intitle:vista etc.. which already pre-sorted for me.
What Could Be Improved: I wish there were more options for setting these things up. I'm not really asking for more boxes, but rather something like they have on the default web-search box "advanced search" where I can maybe run a little wizard to help me optimize my settings. Also, I'd really like to see something like GMail that would allow me to see a past sample of what I might actually receive if I had run X query Y months ago or something to that extent (just as under GMail you can "Test Search" your email)
From Where: Yahoo! Alerts
What Are They: OK, this is such a subjective question, but I'll specify that I do like Yahoo! Alerts more than Google Alerts. I'll get into this more, but I again set this up to give me results on Microsoft.
How Helpful Are They: As I mentioned above, I really like Yahoo! Alerts. In essence, they're very similar to Google's alerts (and without as many options), but I really like how you can setup News under the generic "News" and "Breaking News". I've actually only set up "Breaking News" originally so I'd get everything, but then switched to experimenting like in Google Alerts to only Keyword News with "Vista" title:Microsoft etc.
What Could Be Improved: I'd like to see a little integration between both Google and Yahoo alerts. What I mean is, I'd like to have an RSS feed as such maybe created from my email alerts, then I'd Yahoo Pipe it up and make a feed to take out all repeated stories (which there are alot of). I mean, it's fairly straight forward to see why there'd be overlap, but what's moderately frustrating is the fact that it's currently very difficult to manage these similarities.
From Where: CNN Alerts
What Are They: OK, so the common theme here is to test out for myself what alerts return what. When I set up a general control of the "same" type of query, I tracked "Microsoft" as a whole (don't do this unless you setup a separate email folder), "Vista" and "Xbox".
How Helpful Are They: Ehh, read below for the reason I think these alerts could be more effective. Generally, the way I set these up, they were just a subset of information I *might* get from either Yahoo! or Google Alerts--and as such, weren't as helpful as the first two times (starting to see alot of overlap).. I realize I'm using pretty basic queries, but if I used the same specific queries for each of these alerts, the same thing would happen. If I were to use different queries for each of them, I'd ditch this one all together for limited flexibility.
What Could Be Improved: I'd like to see more flexibility in search queries. As of right now, it only appears that " " and - works for syntax, but clearly this is limited. In general, I still get news that's relevant, but as I said, just doesn't allow for maximum flexibilty.
Posted by asmuench at 08:00 AM | Comments (0)
October 24, 2007
10/23/07 Blog #5: Yahoo! Pipes Tutorial
This is a tutorial that I did on how to use Yahoo! Pipes to create a customized RSS feed that will filter many feeds for the term gphone. Although very simple, if you work through this tutorial, you will be more comfortable building more advanced pipes. Many samples are available on the yahoo site.
Posted by asmuench at 11:18 AM | Comments (0)
September 23, 2007
9/23/07 Blog #4: Searching for RSS Feeds..
OK, when I set out to find RSS Feeds, I wanted to find something related to stock picking. The way that someone in some place decided to choose some stock and make or lose money. Essentially, I was looking for someone's thoughts on the market and why they thought the way that they did. Naturally then, I was looking for my target RSS feeds to come from blogs--so I began with the big name: Google
My adventures with Google's Blog Search
Score: 9/10
Google Blog Search
I must say I was rather alarmed to see such a different interface from the other Google related sites. NOT. This was just the same, just as easy, and returned results that were ehh you'll see. I first searched for "stock picking" and it immediately returned results with stock picking in the title, but nothing really much to do with individual stock picks (except for some guy who thought the world could benefit from his stock picks from two years ago). So I decided it must be my search query and I researched for "my stock picks" remembering that I was searching for blogs. Jackpot. I got a bunch of pretty sweet sites, usually people with only one pick and their articulation on that pick. I also returned Jim Cramer's Mad Money picks (good thing, because if this wasn't in the top 10, I probably would have knocked Google for it).
Next up: Blogdigger's Head-to-Head with Google
Score: 2/10 (for returning something at all)
Blogdigger
First off I thought it was pretty cool how I could search by date or relevance. Although relevance is very important, if it's not dated it's almost useless (in terms of stock picks), so I decided to search like this.
OK, so I thought it would be rather unfair to start with "my stock picks" since we're doing a Google head-to-head, so I went for "stock picking". The results: disappointing. I mean, they weren't great with Google either--but some of the results I got were Fashion, tiles, etc. Not on target. The final result on the first page was actually quite good. But 1/10 isn't something I'd go back to in a heartbeat. Next up: "my stock picks". Nope. Even worse (believe it or not). Haha, you would think it could even atleast return it's last of 10 results from before, but nope. It chose to have completely irrelvant links. Let's try searching by relevance instead: MUCH BETTER. However, I don't care. Stock picks are very dated--so this again is useless information. Hopefully our next tool is a little better.
Next up: Let's try something just for it's name--IceRocket
Score: 1/10 (random results don't count)
IceRocket
I'm not sure if this is some kind of joke blog search, but it's not that funny because I can't find what I want (that's not meant to be funny--that's meant to be very serious). Haha, but in actuality, this search engine put THE MOST RANDOM results in a list and returned them. I feel like instead of searching for "stock picking" or even "my stock picks" I could have put Algerian Horse Racing, and it would have given me similar results. Maybe I'm not IceRocket savvy, but this search engine blasts right out of anything useful into a whole new universe of information I'm not looking for. Just because I don't think 0/10 is nice, I'm adding one. (It's interesting too because any way I can possibly think of to return me something remotely relevant is not possible--boy am I glad I had this one last or I would think I was doing something dead wrong)
Posted by asmuench at 02:46 PM | Comments (0)
September 16, 2007
9/16/07 Blog #3: The Deep Web through Exercises
Ok, this blog will attempt to find and differentiate between different search engines on the internet given the same information. We will be using "Timber Industry California".
#1: Google
A pretty obvious place to start--what do we expect to see? Well I've noticed that since Google is a full text search engine, you'll see both Timber and Industry and California in both the titles and in the descriptions of the search results. If we include it in quotation marks, it will all be together. I think it's kinda interesting how it also bolds CA in the search results since CA = California--smart engine!
#2: Yahoo Directory
Wow. Some damned good results. Instead of just in titles, the Directory is created by people who surf through the pages and essentially "tag" certain key word about a page. These results are very business oriented.. wonder if they get priority when people go through and "tag" the internet. One more thing to note: 155 results!! This is a very small and specific result set.
#3 Yahoo Web
There's really not too much a difference I've noticed between these "general" searches and the Yahoo and Google web searches. Both return a ton of pages, and they're pretty similar in content. The same pages don't necessarily come up--but the results are similar in nature I feel.
#4 Scirus
This is a science specific web search--and it did just that when I entered "timber industry california". I was able to find a few scholarly articles which had results including timber and industry and california and some of which were just timber, some industry etc. Interestingly, alot of these pages probably wouldn't show up until later on a Yahoo or Google--and some at all considering Scirus looks through some databases which Yahoo and Google don't have access to search (or don't know how).
OK--this is pretty cool. All scholarly articles. You don't know the total listing of available journals it's searching, but it's doing it in the same way I feel it that it does general web searches, albeit with far fewer results (less than 30k).
OK, I don't like this one, but that's probably because I'm not really *actually* interested in "timber industry california". Anyway, I was able to select a few databases and search these for our term. What I noticed is that not only will it search the databases for this term, but it will obviously show up more in the business/economics section.
QUESTION: Since I can't access alot of the files directly I'd like to know if it makes sure whatever it searches has the three words: timber industry and california, or if the words actually have to appear together like this? Does anyone know??
This search engine looks for other search engines to complete your search. When we entered "timber industry california" nothing came up. Geez! I would think there would be a search engine out there somewhere that would look for stuff related to the timber industry and in california!? Apparently not.. until next time have fun poking around at these sites.
Posted by asmuench at 04:37 PM | Comments (0)
September 10, 2007
9/10/07 Blog #2: Yahoo's Dual Functionality
(Yeah yeah, it's got more than just dual functionality--but in terms of what I'll be discussing in this blog... hold your horses)
Today in BIT 330, we had a quick exercise assigned at the end of class that essentially compared the Yahoo and Google searches. Without too much of a surprise, Yahoo and Google return different results, but generally the same pages are placed at the top of the same queries.
What really interested me, however was the Yahoo directory. Oddly enough a few sites have this dual way of searching (even the umich site has a directory listing--even though it's a student directory) but I've never really taken the time to understand the features of each search. So I started digging and wanted to find a page that would explain to me exactly the ins and outs of this search.
If you follow me to the page I've linked above, you might be able to better understand the difference. Essentially it boils down to this: there's a "directory" or "network" and then there's the web. Yahoo's directory or network is a subset of the web--but it's categorized (as you would expect of Yahoo--especially after seeing the content heavy front page). This categorization sometimes helps make queries more relevant and precise. Where as you might not receive all of the results you're looking for, the results you'll return will be better matched and will require less searching. !Whew!.
Ok, this might sound a little strange, but what fascinated me most about this "directory" was the fact that there was a starting point for all categorizations. Essentially Yahoo decided that any and everything could be first assigned into 14 categories. No more, no less. This is the starting point--one of these 14 things. Pretty interesting. I wonder if they had a computer program assign all webpages to X number of categories and then sort to reveal the most defining characteristics. Oh yeah, I almost forgot:
Posted by asmuench at 11:24 AM | Comments (0)
September 09, 2007
9/9/07 Blog #1: What do Andrew Muench and BIT 330 have in common? (basically what I'd like to learn in this class)
My computer skills are pretty on par with my numchuck skills: pretty sweet. OK just had to throw that out there--but in actuality, I've realized that my computer internet searching, finding, (attacking) skills aren't what they could be (in other words they ARE on par with my numchuck skills: pretty non-existent)
So I found this nice little class taught by the only professor who's consistently funny--and it hopefully will allow me to complete my other b-school and other related work more efficiently.
That is the goal.
We all know Google isn't the only search engine out there, you've got Yahoo, Hotbot, Baidu if you're Chinese, but I always was under the impression that "one" was better than the others. My goal is to learn a variety of internet searching skills that will draw on the many strengths of the different resources out there and available. Google beta stuff has been pretty cool so far, and they've come a long way in providing other specialized searches--but I know this isn't all it, and Scott Moore knows where to look for these other resources. As such, I will be taking his class with very high expectations. Expectations that I will finish alot of my work quicker and actually have better quality information.
Keep tuned for what I'm learning during the semester.
Posted by asmuench at 11:22 AM | Comments (0)