THE USE OF SEARCH ENGINES IN ARCHITECTURAL RESEARCH PROJECTS
Written an compiled by
BANKOLE-OJO, OLATUNBOSUN
Arc 01 9214
Course lecturer; Prof Ogunsote.O.O.
Date: April 2008
Table of content
Table of content
List of plates
INTRODUCTION
THE EMERGENCE OF SEARCH ENGINES
Free-Text Search Engines
Alta vista
Lycos
Hot Bot
Northern light
Index or Directory based search engines
Multi or Meta Search Engines
Natural Language Search Engine
Resource or Site-Specific Search Engines
RETRIEVING MULTIMEDIA INFORMATION WITH SEARCH ENGINES
HOW SEARCH ENGINES WORK.
CONCLUSION
REFERENCES
Internet references
List of plates
Plate 1; The Alta vista home page3
Plate2; Search results of Alta vista5
Plate 3; Lycos home page6
Plate 4;Search result of Lycos7
Plate 5; hot bot home page7
Plate 6; Advanced search in hot bot8
Plate 7; search result with Northern Light8
1
INTRODUCTION
According to Osuala (1990), a research simply means the process of arriving at a dependable solutions to problems through the planned and system collection, analysis and interpretation of data. It is the most important tool for advancing knowledge, for promoting progress and for enabling man to relate more effectively with his environment, to accomplish his purpose and to resolve his conflicts. However, there have been a lot of various processes involved in writing a research thesis. Part of these processes is the information retrieval stage where information about related subjects and topics are gathered. These information however can be sourced from different medium including printed materials; texts, journals, encyclopedias, dictionaries, press releases. Non- printed materials; such as audio-visual sources like tapes, slide shows etc. The advent of the internet[i] as a result of globalization makes the possibility and the accessibility of information retrieval far easier than any of the above listed sources as this can be accessed even in ones own private bedrooms instead of going to public libraries for referencing printed materials for research purposes. The important change for the information professionals as the internet is not interested in geographical location and the information needed can be found regardless of the location in the world. Consequently, global information is available on the local desktop. When using the internet, contacting the appropriate source for the information needed becomes so essential and this marks the main area where the internet search engines finds their application.
However an effective searcher will be able to blend all these different resources information got through the search engines into a single collection, using whichever element which is best to answer the query[ii].
THE EMERGENCE OF SEARCH ENGINES
In the early days of the internet it was reasonably easy to find information of data filesusing a variety of soft wares that are usually command driven; ie there is a prompt to type in the command to be executed rather than using a graphical interface. However, with the proliferation of the data brought about by the growth of the web, these systems with such names as Achie, gopher and the veronica became increasingly unable to cope, in other to overcome the lack of retrieval facilities, anumber of organizations and individuals began to create their own searching and retrieving information. (Phil. B 2002).
However, the problems associated with the early form of search engines are;
- Despite their usefulness, very few of these search engines had any kind of library or information background.
- Although they were (and still are) doing a sophisticated job of finding and indexingall those information, the early implementation was quite crude; they were not designed for trained searchers, but for people who had never done a literature search in their lives, and as a result they did not make any use of Boolean operators, proximity searching, wild cards, truncations or any other thing that are taken for granted..
- Some of these search engines have hardly moved forward at all, and still the only way of finding appropriate web pages is to throw as many search terms at them as possible.
- They return all the data they can return indiscriminately not being able to make any judgment on the value or the authority of the information they find.
A search engine is used to impose some controls over the web (which is large and growing at a tremendous speed under its own anarchic mass but strangely enough, in the main going, in one direction) and to put some structures in place. It is of great assistance in allowing us to quickly find and exact piece of data required some in less time it takes to articulate the query.
However, there are basically five different types of search engines available for research and enquiry purposes and they include;
Free-text search engines
Index or directory-based search engines
Multi or meta-search engines
Natural language search engines
Resource or site-specific search engines
Free-Text Search Engines
This is used to search for any single word or a number of words and in some cases a phrase. There’s limit to the choice or patterns followed to search for example the name of an architectural firm, a line of poetry, an architect’s design philosophy, an architect’s name, a foreign language term, just about anything in any format. Free text search engines are very useful if the idea of what is being searched for or when looking for a concept which can be defined in a small number of words. They are less useful when the search requires a broad overview of the topic (query) or when the search is being carried out in an area the researcher doesn’t know very well and consequently have no idea as to the best terms to use.
The characteristics of free search engines are
They will accept any term the user wishes to search for
It can search for terms in any combination and arrangement
It can search for phrases as well as single words
It allows users considerable flexibility in choosing how to search.
The different types of free-text search engines are
Alta vista
This is one of the best known and the oldest internet search engines available (the other being yahoo) and was launched on the 15 Dec.1995 at The aim of the Alta vista service is to index the entire world wide web and the Alta vista team has estimated to have indexed over 350,000,000 web pages (Phil.B 2002) making it one of the largest search engines.
Plate 1; The Alta vista home page
Using the Altavista search engine
Alta vista will automatically or searches together unless you tell it otherwise. This inevitably lead to large result, however it will rank the web pages it retrieves for relevance according to its own algorithms by displaying what it thinks are the best possible matches first, with the less relevant ones father down the page.
A better search strategy however is to include as many different terms as possible, since one of the ways in which it works is to give the highest relevance to web pages which include all the researcher’s terms. Consequently, if the user is interested in the search for post-modern architects, it will make sense to do a search for “post modern architect architects building buildings city cities” and so on. In this case however, the use of astererisk ‘*’ could be used to save some time as it forces the search engine to look for any word that starts with the letters chosen and any other letters will appear after the ‘*’. There are many other techniques used or offered by Alta vista in narrowing down the search results to the very useful ones required these include the use of
+ or – symbols which helps to depict those things that are to be included and excluded respectively from the search but this is rather cumbersome
The quote mark (“ “) is also used to indicate that we wish to search for a particular phrase. This gives a shorter, precise and neater results than the + and – method or approach. For instance let’s consider the fact that the researcher wants to search for ‘modern architects’, using the first approach above, the query can be typed in this format;
+modern architect architects building buildings city cities-architraves-architecture-archifect-history.
From the above query, the excluded words are necessary in order to make the search engine concentrate on the required search as it is very likely it a lot of irrelevances from the search of these words. This doesn’t totally stop irrelevances but at least reduces it. However this is still a very cumbersome query, a rather precise one could be achieved by mere using the quotation mark method which will be in the form of “modern architects’. This is short and precise and forces the search engine to all issues relating to the topic and neglecting all other irrelevances. This is rather more precise than the former as the number of irrelevances given in the search result would have been sieved to the bearest minimum having a smaller appropriate set of results.
Another result-trimming technique in Alta vista is the provision of a number of switches which can be used in combination with single terms or phrase searches to further sharpen and narrow the search down. This is in the format; <option>:<search string> e.g. a query can be formulated in the following format
‘+ “Nigerian architects” + domain: Nigeria’
This search result for this query will be Nigerian architects in Nigeria and non on those practicing outside Nigeria. A further streamlined version could also be
‘+ url: Nigerian architects + domain: Nigeria’.
Also the use of capital letters in search queries for Alta vista also helps limit the variability of the search result. This helps distinguishes between the search for a noun or verb or other words in a sentence. For example , the search for the name Oscar Niemeyer could be got by merely typing “Oscar” as the query but a small letter ‘o’ might prompt the search engine to be looking for every other thing named oscar other than a name like oscar award which still makes the result displayed so cumbersome.
Also, Alter vista sometimes provides a list of related searches which may sometimes prove to be a better search than the one originally ran.
The format through which the result of the search is displayed in Alta vista is as follows
The fist line is the page title as defined by the web page author by the url, then a brief summary of the page which is taken either from the first few words that Alta vista finds or from a special meta tag used when designing the page, under this is the option to translate into other languages and after this is the related pages, an option to view more pages from a particular site.
Plate2; Search results of Alta vista
Advanced search in Alter vista
Theuse of the advanced search option is important once one become familiar with the general search process or when there’s a specific query in mind. Most searches can be run perfectly well using the simple general form, but the major disadvantage of it is thatAlta vista controls the relevance ranking process but in the advanced search format, you have more controls on how words or phrases are ranked. Boolean operators can be used to greater effect and the user has the control on the data field. However, Boolean operators can be used when all other technique doesn’t work are they are usually written in upper case. Here, Alta vista gives alternative however and allows the user to use the characters & for AND, | for OR, ! for NOT and for NEAR.
Lycos
Established in May 1994 at It is one of the oldest search engines and began as a project at the CarnegieMellonUniversity. It is used to be just a straight forward free text search engine allowing the user to input any term desired in order to locate appropriate websites. But in common with other free-text search engines, it now also provides a directory listing of sites in a hierarchy. The topics in the Lycos hierarchy operates in a similar fashion to the directory outlined in Alta vista
Plate 3; Lycos home page
Lycos supports a number of Boolean operators which can be incorporated into searches and these include the following;
ADJ- Modern ADJ Buildings retrieves site which contains both words next to each other in any order
NEAR- Building NEAR Modern, retrieves sites where both terms are within 25 words of each other in any order.
FAR-Modern FAR buildings retrieve sites containing both words more than 25 words apart.
BEFORE- Modern BEFORE building is the same as the AND operator but the first word must come before the second word at any distance apart.
It is also possible to specify the order of words by specifying O( meaning Order) immediately before the operator.
OADJ-Modern OADJ building is equivalent to searching for “Modern building”
ONEAR-Modern ONEAR building retrieves sites where the terms are within 25 words to each other and ‘Modern’ comes first.
OFAR- this works like ONEAR except that the terms must bemore than 25 words apart.
Plate 4;Search result of Lycos
Hot Bot
This is a search engine that was created by the wired ventures inc. the same organization that publishes wired magazines. It can be found at In common with other search engines, Hot Bot provides both easy and advanced interfaces. At Hot Bot, you can also specify if the page should contain an image, MP3 files videos or PDF files. Hot Bot also provides a “house by category option” which has the look and feel of the Alta vista approach. Hot Bot has links to search for other types of information such as Usenet, yellow and white pages (US only), e-mail address, domain names, classified events, top news site, homes and loans, stocks and shareware. The advanced search facility is very flexible with considerable functionality.
Plate 5; hot bot home page
Plate 6; Advanced search in hot bot
Northern light
Northern light was established in 1995 ( while not a new search engine, has over a very short period of time become one of the major player in the search engine market. It has grown consistently since establishment, currently indexing315million web pages. The simple search interface allows the user to input single terms or a number of terms to do phrase sentencing and to use Boolean operators. It also supports the use of right-side truncation inside a word (using the asterisk) or single letter replacement using the percentage symbol. Consequently, the user can search for:
Arch*- which will result in pages that contain words that begin with “arch” such as architects, architecture, architraves etc
Philo*phy- which will return sites containing words beginning with “philo” and ending with “phy”, e.g philosophy, philography.
Gene % logy- which will return sites containing words beginning with “Gene” and ending with “logy” separated by a single letter e.g Genealogy and Geneacology. This can be useful in commonly misspelled words.
Plate 7; search result with Northern Light
Other free-text search engines
There is almost no limit to the number of free-text search engines available but some other existing ones apart from those discussed above are;
AOL anywhere at
ESporting at
Euro Seek at
Google at
Look Smart at
Search Europe at
Web Crawler at
Mamma at
Index or Directory based search engines
This search engines take a rather different approach to providing the user with information on the site to visit. Their emphasis is on clarifying information under series of major subject headings and then subdividing this information into a tree-structure of more specific headings. Sites are listed as appropriate in this directory structure which is similar to the library classification scheme. The subject headings and the subheadings can be used to guide the user through the vast mass of information until they are able to locate exactly the right section which covers the interested areas. An in-depth knowledge of the subject is not required since the user can not stop drilling down through the tree at any point. The most famous of the search engines of this category are the
Yahoo[iii]at
Magellan at
Net find at
Global Online Directory G.O.D at
Excite at
Multi or Meta Search Engines
A multi-search engine doesn’t actually search anything itself; instead, it takes a query and passes it on to a selected group of search engines. Once the result start coming from these individual search engines, a multi-search engine displays the results on the screen. The more advanced engines will collate the result, removing duplicates and put them into some sort of sensible order. Multi-search engines are useful when trying to obtain a comparative listing of websites that cover a particular subject. Using a multi-search engine implies that the user only have to visit one page and all results are brought back to the screen, therefore limiting the amount of work to be done. However, the disadvantage is that the multi-search engine uses syntax that must be configured to be able to work with the single different search engines which means more work in another direction.
Another type of multi-search engines simply provides the link to lots of other search engines without giving the user the opportunity to input search terms or queries to be forwarded to them.