module 4: searching the web.

This is a timely task for me with the concepts assignment due date fast approaching. My skills in searching for quality information on the internet are very basic and i have tended to waste a lot of time so far without much thought before hand as to how i search . I just think Google. But doesn't everybody? Google is so popular that it has become a verb. when you need to do a search, you go and "Google it."

This task requires a meta search engine, the suggested ones being Copernic or Sherlock. Well i am a Linux user so both of these programs are useless to me. What to do? I did a Google search of course! I searched for meta search engines and came up with some good hits. Websearchguide and the Berkeley Library website i both found very useful.
This is what i did.I searched on Google for the term "instant messaging" and got 37,300,000 hits. The first hit being at Wikipedia on the subject of instant messaging.
Next, i searched using a meta-search engine called Allplus to compare results. Allplus works by delegating the search query to other search engines. In this case, Google, msn, Ask and Yahoo. Allplus also searches news sites,images, video and Twitter would you believe. Using the term "instant messaging", Allplus came up with a combined number of hits of 197,599,000 hits. Wow!
So how should i interpret this? Theres lots of overlap and duplication in the search results, but there would be lots of hits unique to a particular engine too.
I also tried using the meta-search engine, Surfwax. which the Berkeley library site recommends for serious deep web searching. I tried to register as a user so i could take advantage of setting up some custom searches, but the registration process was closed. From searching Surfwax. i did get lots of results but only from a 5 or 6 sources.
My experience of using both Google and using a meta search engine is that they just give too much information, much of it irrelevant and unhelpful. How are you supposed to process so much. Much better to get only a handful of hits that are useful.

Regards the issue of searching the "invisible web" or "deep web",it is not as straight forward as it appears. From my understanding, there is no one particular search engine that will perform this task adequately although Copernic comes close from some of the reviews i read. But the issue here is that a huge amount of the content on this part of the web is contained in databases, the content of which, may not be generally accessible to to the casual user of a search engine, even a meta-search engine. I found some of the articles at Websearchgguide on the topic of the invisible web or deep- web very helpful. They suggested i use the term, "database" together with the topic of interest in my search query. When i found a suitable database, i could drill down through that to find suitable information. I tried this just using Google. I used the terms: "Computer mediated communication" AND "text based chat"AND database"
and found some relevant databases to search.
They also suggested i try CompletePlanet, a site that allows you to access roughly 70000 searchable databases. If I could find a suitable database then i could search within that. I used the term "instant messaging" and came up with a list of 80 databases. I was then able to narrow my search to databases within the subject of social sciences which gave me five hits, one of which was titled "journal of computer mediated communication " and looked very promising. Of course i have now added CompletePlanet to my list of search engines for future use. Actually CompletePlanet is a very impressive website.

Boolean Searching on the internet.

To get the biggest number of hits relating for a search term, you need to use synonyms. This casts a wider net over the possibilities. You do this by using the Boolean operator: OR . For, example, for the term"invisible web" you could construct a search thus: "invisible web" OR "deep web" OR "hidden web".
When i did this using Google, i came up with a massive 42,200,000 hits, substantially more than just using one term. Thats just part of the story though. What do i do with so many hits? Its not quantity i want, its quality and to do this, i have to try another search strategy.

To get the most relevant information according to what you searched for, it is crucial that you use many keywords to qualify your search term. You do this by using the AND operator.This means that keywords on either side of AND must be in the document.This will narrow your search considerably to a manageable number of sites to investigate which hopefully are relevant. For example, i used the following search query in Google: "text based chat"AND databases AND"instant messaging" AND "social impact"AND "psychology".
This narrowed my search down to 11 hits and most of them looked pretty promising.
It is easy to limit your search results to those simply from university sources by narrowing your search results to the .edu domain. This is easily done through the Advanced search preferences. Also, when you use the term database in your search, you notice that a lot of the results are indeed from academic institutions.

Organising search information task

Choose the three best sources found (so far)

1. http://www-usr.rider.edu/~suler/psycyber/texttalk.ht
2. http://www.livescience.com/technology/050301_internet_language.html
3. http://mastersofmedia.hum.uva.nl/2007/04/11/text-chatting-versus-face-to-face-talking/

I have used Zotero , a Firefox browser extension to collect manage and cite my reference sources. Below is a screen shot displaying bibliographic information for the first of my sources. I think if you click on the image, you will get a better view.


Zotero opens in three panes. The left pane contains folders called collections. Each collection is connected to a topic. My sources so far are all gathered together in the middle pane and are contained in a collection (left pane) which i have called "text based chat".This helps keep all my sources relevant to my topic in one place. The pane on the right contains the bibliographic information related to a particular source. Most of it is captured automatically,but it can be edited. I have created an abstract made up of key words that help me remember what the article is about. it also contains the URL, author and a link to the original website. So for each source, i have created an abstract and stored the bibliographic information.
Zotero also creates a snapshot of the webpage which is stored locally. It then lets you annotate to your hearts content. In other words highlighting and adding comments much like i might do in the margins of a book or article. (see below).



Finally, this application lets me print out my accumulated bibliographic information in a number of formats including APA.