Abstract: This paper investigates a few strategies for seeking information on the WWW. The strategies are: Seeking information with thematic directories like Yahoo, Dmoz, Loosmart; use Search Engines like Google, AllTheWeb, AltaVista, etc., the using meta- Search Engines – Metacrawler, One2Seek, Mamma, etc., the using a few techniques for a limitation or a expansion of results ; for using URLs, portals ,etc.
 
 

Keywords:Strategies, Information Search, Information Retrival, Search Engines, Meta- Search Engines, Portals, Intelligent Agents

There is ceaseless increase of the Global Internet services. One of the most used options for information search is WWW. Actually World Wide Web (WWW) arose in 1994 though Tim Berners-Lee from Laboratory of Physics, Geneve suggested hypertext system and access to documents in the Net using hyperlincs in1989. World wide web pages continuously increase

their number. There are more than 2 billion and seventy million pages nowadays. Specialists created and continue to create and improve instruments for fast finding of documents – directories, general and specialized, regional, national and meta searching machines. This research is devoted to information retrieval in Web and strategies for its realisation.

After booming development of the data base systems and systems for office automation during seventies and eighties of twentieth century, the arising and development of “information retrieval systems” is one of the most progressive directions for the information systems. “Information Retrieval” is one of the most successful areas of the natural languages processing.

According of the British computer Society Information Retrieval is purposed to ensure fast, effective and productive methods for describing, managing, searching, return and presenting of the information from CD and Internet.

Immediately after sixties were created information retrieval systems on request in text data bases of certain language. One of the founders of this direction is the Dutchman Keith van Rijsbergen.. He is working on the field of information retrieval since 1969. The group for “Information Retrieval” led by him has powerful research program based on the theory as on the experience. The purpose is to give new not used to this moment and higly effective methods for access to multimedia world. They develop models applying logistics, contingency theory, computerised lingustics etc.

I. Directories, search machines and meta-searching machines

One could differ between directories and searching machines. The dirrectories are arranged by specialists. The information there is classified in the form of structural tree. Changing of the page or pages does not reflect automatically on the redactor’s list. Open Directory Project, known as Dmoz, is Netscape property and it was realized by volunteering editors.

The Open directory on http://www.dmoz.org is just a data base.

Partners, who use the Open Directory, including AltaVista, Netscape, Lycos, HotBot etc. often posess many other data bases.

The open directory Dmoz is characterised by one of the largest data bases.

It is newer than the open directory of Yahoo. Lately it characterises with changed interface for searching the information in Web or News groups. It lets also searching of FTP resources.

Search Engines are programs that let the user to search for information by key words. They create index of data bases or Internet sites on the ground of headline, key words, or full text. They suggest an interface to the user for submitting the search request. After finding the result, they generate a list of findings and let the user to click on certain hyper link. The efforts of the specialists are directed to improvement of the interface – i.e. creating of various options for the different groups of users among which in their native or preferred language.

The searching engines consist of three parts – Crawler , Spider , Robot, who travels constantly over the Web space. While coming back to pages the spider constantly search for changes in their content. The first spider is created in 1993. It is named World Wide Worm. All the spider finds goes to the second part – the catalogue. The catalogue is as a giant book. It is called also index. The third part of the Search Engine is a program, which filters millions of pages, to find the most relevant to the request. It arranges the results found.

There are also hybrid Search engines, which combine the both possibilities – associated directory and robot (for e.g. Yahoo, which belongs rather to editor directories, but submits access also to sites that are found by Google Search Engine).

Searching engines give as a result a file of hyper links to web sites, which contain the key word or phrase. According to some researchers of the retrieval process – for e.g. the site http://www.wgi.com/internet.htm, it is asserted that the No 1 way to find something in the Web is to use Search Engines. Four of every five users report that they use Search Engines daily.

According the data of the users, 80 % of Search Engines stop the search of the sites after 30 results. To answer that assertion we would announde that there are search engines which find and show much more sites. Among them they are AltaVista, Northern Light, Google, Allthe web. For many requests for search they produce huge number of sites. So for the specific phrase “Search Engines Optimization” AltaVIsta finds and shows 54 results. Many of the specialists are now directed to developments that are connected with optimisation of Search Engines and the information

security. Various products are created such as Search Engines Commander (a pack with revolutionary features, which lets the user to control Search Engines, to use templates, smart key words, replace etc. It is working on the back word mode).

Meta searching engines do not support data bases with addresses of the sites. They support data bases of searching engines. When using meta searching engines, the search is accomplished in parallel with several basic searching machines and directories. In some cases, near the results the means of search is cited. Some search machines for meta search

of the information are developed to the level of capacity to find doubled page addresses, obtained as a result of different engines of search and directories..

Some search machines the means used is indicated near the result of search. Some search machines for meta search are developed to such level of capacity and they can find out the doubled addresses of the pages, obtained as a result of various search machines. Among them are: Copernic (which can be downloaded for free), One2Seek, which gives absolutely relevant and non repeated results, Мetacrawler etc.

II. Strategies for information retrieval

According some authors there are three basic strategies for search:

1. Use of reliable or approximate address – URL.

1. Use of thematic directories

1. Use of Search Engines

According to D. Amor the fastest way to find something in the web is using of meta search machine. He gives his prefer to Metacrawler.com.

The site on address: defines the criteria for assessing and research of the existing meta “search engines”.

Search of headlines and URL addresses.

If one enter in Excite U:"Information Systems" OR "Information Technology" OR "Informatics", it can find 2103 sites, which include at least one of the expressions which is included in brackets.

According to our opinion all information retrieval systems in the Web could be classified in the following way:

II.1. Use of the directories and Search Engines
The basic classification of the ways for information search includes searching of the information by key words and expressions, option for thematic search (by categories) and combined search. If the subject to be searched does not exists as a category or sub category, the best thing is to choose respective category or sub category or to make combined search.
Forwarding of an enquiry for searching by more than one key word or use of an expression containing logical operator restricts the search and increases the relevancy of the findings.

On http://www.ibscorporation.com/search_engines.htm one can see an interesting classification of Search Engines.

Every strategy is realised by various approaches. They may be used also for comparing of the possibilities of the separate means of information search in the Web.

Information retrieval through directories

Looksmart directory suggest an option for search in 10 categories and 8 directories such as United Kingdom, Canada, Hong Kong etc. Certain interesting directory is Work & Money. If its sub directory Industries is selected, then in a consequence one may reach sites of the companies that produce specific products – for example -

–Industries>Manufacturing>Food & Beverage>Food Companies>Dairy Products>Cheese Companiesґ. Looksmart suggests also Looksmart Centres – Careers, Real Estate – all 15. The center for insurance is also interesting - with information for all sorts of insurance. Even pictures are available for search. So for “Imae” it finds 2005 objects. Webcrawler also submits very rich information due to the help of its partnership with Looksmart. Webcrawler offers more than 1.5 million web sites and 100 000 organised categories of the sites in its own directory.

Yahoo contains 14 categories. With categorial search one can view interesting sites according the respective themes and sub themes. So for Regional>Regions>Europe>Government one can see the sits of international organisations, government etc. Global symbols are admitted to be used. If the search is by the first symbol* and there is inserted a global symbol after it, then interesting results can be obtained.

118654 categories and 820577 sites for S*.

Interesting sites in this query may be found in Home>Science, Finance etc.

It permits an access to 3500 companies accounts for free. There are 30 categories and 2361 sites found for “Quotes”.

There are 135 resources found for the expression “Analysis for Portfolio”+free. Respectively for the expression “Analysis for Portfolio” AND “free” Hot bot finds 1000 sites. For the last expression Northern Light finds 364 positions in 196 soruces. With Yahoo one can find out interesting resources on the field of students’ working with computer system, with Windows”98 etc. . For the expression image(*.jpj OR *.gif) Yahoo finds 129 thousand web pages.

The directory NBCI (http://www.nbci.com) suggest an access to 20 group categories. Graphs and indexes for Stock market exchange are available for browsing. NBCi admits use of global symbols in searching by key words, permits sensible search, restricts the search by language and date.

One of the newest directories is this of http://www.hyperseek.com/.

Search with demoz –http://www.demoz.com.

It is possible to find easy for trademark. The user should determine what type of trademark is searched – Federal, Canadian, European. The site suggests also web hosting service only for 14.95 $ monthly. They offer the pack SmartSuite Web Hosting, which is appropriate for everyone who starts electronic commerce. The site suggests to start http://www.register.com/businessresurce/, where one can trace for example 14 steps for building of a successful website.

Dmoz.org, which data base includes 2 million and 700 thousand sites. more than 387 thousand categories organised in more than 16 group categories in 64 languages, if one chooses World->Bulgarian, gives

an access to 875 Bulgarian sites may be obtained which are grouped in 16 themes. So if one choose education, then they could be available to be browsed.

Certain interest represents the way Search Engines arrange the results found after specific query. Usually on the top of the list the most relevant results are put. The agents follow rules the most important of which are the place and frequency of the findings of the keyword in the site.

Indexing machines identify all pages, which contain the search terms of the query, while “search engines” identify small part of these data. Indexing machines ignore hyper links while some search machines as “Page and Brins search engine” base the search on Internet hypertext structure.

Indexing machines first identify pages which contain the users terms for search and then they apply a function of arranging the pages in order to

identify the relevant pages. Various indexes use various characteristic features as for example frequency of the findings, preferences for the text of the headline, presence of meta tags, statistics of the previous visits, relevancy for the search terms. The context of the terms in the query also could be included, but this requires involvement of the user.

There are intelligent agents are used too in order to support the Internet search. Meta- search engines combine the results of the index machines, through combining of their ranking functions. In order to show the importance of the information contained in , except of citations, they use hyper links. They are used as a references, recommendation to certain site.

The authors include hyper links to their sites. The new searching machines use hyper links to identify the collection of the connected documents under the given topic. Google expands the use of the hyper link structure relying on the text for identifying of the relevant pages.

Search Engines check if the key words appear in the headline or on the top of the text as first paragraph. They accept the fact that the page is relevant, when the key words are at the beginning. They also analyse how many times the key words appear in respect of other words and in case of the more often appearance they increase the relevancy.

The Length of the meta tags is another criteria which plays important role when arranging the site in the result. The recommendable length is 150 symbols, where some search machines as HotBot accept length only to 250 symbols. Some search machines increase a little the relevancy of certain sites due to advertising motivation. Exite for example uses the popularity of the links as a part of the method of arranging. Infoseek and HotBot also increase a little of the relevancy of the sites which contain key words in the meta-tags. But Lycos does not read meta tags at all. There are many examples with this searching machine when the sites are highly ranked without meta tags. Search Engines could decrease the level of ranking of the site if it determines the Search Engines as a slam. According to researchers on http://www.make-it-online.com/sf.html some Search Engines as Excite, give priority to the sites with key words in their title tag and also with key words which are repeated often within the whole content. In the same research it is pointed out that AltaVista ranks better. Some search engines and directories can arrange the results according the date– AltaVista, HotBot, iWON, MSN Search, NBCi, Northern Light, Yahoo. Some make also clustering of the sites– AltaVista, HotBot, iWON, MSN Search, Northern Light, Yahoo.

Search with use of Advanced Search
Almost all of the basic directories and Search Engines suggest Advanced or Power Search, where the user enter his query in a special form.
The interface in most of the cases is easy for the user.
By Yahoo one may find for "informatics" OR "Computers Architecture" with Advanced Search 11 categories and 215 sites. Again Yahoo with Advanced Search for “Quotes Insurance” gives 450 web pages.
Excite divides the sites of different languages in its own data bases. It gives better options for search by other languages by Advanced Search. While for the expression “Travail йtudiants" OR "travail pur d'йtudiants"-.ca by usual search there is no one site to be found with Excite unless one use the complex search – then a list of 484 sites is obtained.
Through this way of search the information in the complex expression divides into three parts– for the first part the option is included that it is good to contain, for the second that , it must contain for .ca that all sites which has the fourth part in their address should be excluded. (Canada).
Besides a restriction is introduced for language of the search – French.
Other ways of using of the options are available too. Among the sites found there are also Belgium sites.

II.2 Information retrieval by meta-search machines
According of some researchers the meta-search machines take less time for searching information in every data base as they search in parallel in several data bases of basic Search Engines or directories. In some cases they give only 10 % of the content of the data base, but this in most of the cases is enough to satisfy the surfing user. Some researchers recommend if the results obtained are too much then first to view these sites that are obtained by AltaVista, Northern Light, InfoSeek. Recommended metasearch machines are: WebCrawler, Thunderstone, DirectHit, WhatUseek. , Savvysearch.com, Metager.de, DogPile.
Among the best meta search engines, which find unic paes is also
Ixquick (http://www.Ixquick.com), which is capable to translate the query for search to other languages.. So it gives for “Portals" OR "Vortals" an announcement for 47 unic top-ten sites. Meta search engine Ixquick
Shows near every site with several asterisks the relevancy of the site.
So for our query for portals or vertical portals, it put on the highest place (five stars) the site : http://www.tradeworlds.com/.
Meta search engine Mamma.com finds 16 highly relevant sites for “Information Systems" OR "Information Technology" OR "Informatics". It shows the first 10, before every address in brackets which search engine is used to find the given site. For “arbait” they are 16 sites found.
The use of every meta search engine depends on the following factors:
To which search engines it sends the queries of the users, of the number of the basic search engines and directories, of the fact if the user can select the right word, which Boolein operator they imply, which syntax they use; how to present the results – if they are unite in a common list or they are presented in separate. Meta search machines are useful at search of scientific, terms, phrases in brackets etc.
Metacrawler (http://www.metacrwler.com) gives access to 18 groups of categories. Metacrawler suggests information for search of busines information – for companies, for indexes, for the ways of calculating of the indexes, for the rates of shares of the companies, leading media companies as Forbes, Fortune etc.
If during the category search one choose Insurance Quotes, and on the next level – Auto Insurance, it is possible to choose hyper link to 53 companies in this sphere..
Among meta- search engines highly appreciated by many specialists is CNET Search.com, which gives an access to 800 search engines and its biggest advantage is the access to the channels – using a specialised sites.
Meta- search engine One2Seek (http://www.one2seek.com) is capable to do three kinds of search: - simple, basic and complex.

IV.Seeking criteria

А) Retrieval with use of global symbols.
This means that the root of the word is searched for and in this case there are more sites in various languages. Yahoo admits use of global symbols. If the search is conducted according to the given first symbol and a global symbol put after it, interesting results are obtained. 118654 categories and 820577 sites for S*. Interesting sites for this query are found in Home>Science, Finance etc. The access to the accounts of 3500 companies is free. For “Quotes” it finds 30 and 2361 sites. AltaVista also admits it. Google, however in its aspiration to satisfy quite precisely the queries expels it.

B) Natural language information retrieval.
Google finds about 500 sites for assigned at random text query in English Latest Multi-Lingual Information Retrieval Projects.

C) Sensible search
AltaVista makes sensitive search i.e. gives various results depend on the size of the symbols in the query for search.

D) Stop the search – i.e. expel some words from the query. Such approach is used by Google – it searches not for conjunctions, prepositions and other often used words. So for the query for “Resources for Econoists” announcement is shown that says that “for” is very common word and is expelled from the present search.” For this query Google gives about 402 results for 0.14 sec. Contemporary search engines have convenient for the user interface. When the information retrieval is by key words, they apply technics which broaden or restrict the search.” If the results are not enough then they can remove some of the specific terms from the query. When the results are large in number or just the most relevant documents are searched for, then the following restrictions could be applied:

• Use of the sign – (minus) for expelling of the sites which contain certain terms that are recorded after the symbol.

Use of the Boolen operator AND between key words– for search only of the resources which include also the recorded key words before the operator and these put after it. The restriction for search of the certain domain also permits to search among smaller number of sources.

The restriction for search by language also reduces the number of findings. АltaVista and Lycos for example permit search in 25 languages.
• The use of all fields of complex search in most of the search engines permits also to search in the headline and in address which also can reduce the results.
• The click on some of the folders which are on the left of the screen of Northen Light also restricts the search. So if for PUB:"Information management" there are 569 findings, the click on the folder Electronic Data Interchange leads to the reducing their number to 8.
• Boolen operators can be applied to search by headline of the site.

According to the site http://www.kcpl.lib.mo.us/search/ the most used search engines and direrctories are AltaVista, Northern Light and these given in the table below. There is useful information for directories and index engines in some university sites. On site
http://www.kcpl.lib.mo.us/search/chart.htm one can browse the newest data for the size of the data base of the most of the directories and index machines.
According this data Google has 1,25 billion registered sites. Other sources show that 625 millions are cited as partially indexed sites. According the data from http://www.showdown.com Google is on the first place, and Fast (represented by All the Web) on the second, MSN Search – is on the
Third place, AltaVista – on the fourth, Northern Light – fourth iWon – fifth place.
 

Criteria for making compare	AltaVista	Yahoo	Excite	NorthernLight		Lycos		HotBot ( Slurp –the Web Crowel are used)
Size of the data base	550 mill		250 mill sits	330 mill		500 mill		110 mill
Logical operator	or,and,and not, near	Or, and, and not	AND , OR,AND NOT in capital letters	And,or,not	and,or,adj,,far,before		AND, OR, NOT
By default	Or	OR	OR	And	and		Permits modification of the search terms Through “drop-down list”
If it permits + and -	Yes	Yes	Yes	Yes	Yes		Yes
If it permits global symbols	To require at least 4 symbols before them and also it is permitted to include up to 5 symbols additionally which are not permitted in the beginning.	Yes	No, it searches by default automatically also for plural	Yes * %	Yes, *		Permits at the beginning and at the end of the word.
Type of search	Simple, Advanced	Simple, Advanced	Simple, Advanced	8 types of search – InvestText Search, Search News, Search Quotes, Geo Search	Simple, Advanced.  Search in the headline or in Url is permitted     .		Simple, Advanced. Advanced is capable to search pages which contain Java, JavaScript,VRLM(3D), Acrobat etc.
What kind of restrictions it uses in advanced search	According the date, language, Boolen expressions, relevantness	according the date,where to search and among Yahoo or Web sites	Domay, language	С “Power Search” – 10 types – language – 5 languages, country, where to find –  in which type of sites, according the headline ,name, publication,dates etc.	Permits search in 25 languages		According the period, language -  10 languages . Advanced  Offers options for search of pictures, audio, MP3 and sites which contain modern technologies such as Java, JavaScript, ActiveX etc..

All means for information retrieval in the list index media objects and search for pictures, MP3, video and also links to sites. Еxcite finds for “Economics Informatics” OR “economic informatics” 761 sites. Google respectively finds for “Engines Features” 421 sites for 1.84 sec. Northern Light finds for Usenet (without quotation mark) 985,610 findings.

This is the only search engine which suggests the option for search In folders, where the findings are arranged according to certain criteria.– Northern Light offers interesting options for search of business information. So if one browse the headlines of some of the folders it could be seen that the discussion groups are formed on various subjects, but it is possible to say that they are builded basically on two principles – interests on the field of profession and professional searches, interests in entertainment area – mainly discussion groups of fans of singers and actors.
An interesting site is - http://www.cyberfiber.com/index.html, where hyper links of newsgroups are shown in alphabetical order. It is possible to select a hyper link to Arts and Literature newsgroup and Business and Marketplace, Online Services (USENET), Education, Science and technology etc.
Selecting Groups on the basic screen of Google one may find a list of thematic arranged groups. If Software is inserted in the search field, without quotation marks, a list of sites of discussion groups appears about 20,700 millions. (only for 0.29 sec. Here are some more addresses – of the site of International Institute for software tests – Toronto, Canada.
Other addresses could be selected on the site too. So on address:

• http://www.softdim.com/iist/ one may find The Journal of Software Testing Professionals). International Institute for Software Testing is a sponsor of the Journal. On www.jobsearch.co.uk one can find suggestions for job vacancies on software field in England.
•         Yahoo gives for Usenet an announcement that the findings include 747 categories and 3147 sites. Other sites which are structured  in     4 groups hyper links to sites of discussion groups could be seen on:

II.3 Strategies with use of cross-language search and natural language search
The researches on the field of natural language question-answering began in the 60-ties of the twentieth century. The researchers on the field of Artifical Intelligence wonder why the systems for finding documents are restricted to finding documents, which the uer should read in orde to get the answer of the question in interest. Researchers such as Terry Winograd, Daniel Bobrow et all began to build systems, where part of the answers are contained in the data base together with the questions. The idea used by engines for cross language search is to prevent people who don’t use certain language from being harmed. In this respect EuroSearch, a regional search engine overcomes the restrictions for the necessity of forwarding the query and check of the results in language which is not accessible for the user. This enine is designed as a project which unites the efforts of national search engines and suggests multilingual working surrounding.
Every national site is supported by the respective country. National search engines of Italy, Switzerland and Spain are participants in the project. Queries in Bulgarian are acceptable as well. Search engines for cross language search perform more complex functions than these with multilingual search.

II.4. Portal and meta sites information retrieval strategies

For the query "Portals" AND (.de OR .fr) Northern Light finds 666 159 items. Northern Light searches also in a special collection of documents of editors. Information about the nature of the business portals and how to use them could be seen on http://ce.fis.utoronto.ca/tig/
The creators of this site give the following formula: “Portal = contnent + community”. The basic idea is to unite services such as chat, discussion groups, directories of the companies and free e-mail with information content. There are already two generations of portals. The first of them such as AOL (http://www.aol.com), CompuServe (http://www.compuserve.com), http://www.t-online.de etc offered their services only to a certain category of subscribers. Contamporary portals are accessible for everyone. AOL has already 6-th version and offers an options to users from 10 countries – among them–UK, France, Germany, Brazil, Mexico etc. All portal servicies in French are available on AOL.fr. Besides an interesting content, the sites which play therole of portals need also a search engine. One of the most popular portals are these of the companies which posses browsers – Microsoft и Netscape. An interesting Russian portal is Business.ru. Information for the history and development of Russian portals is published on - http://history.ricn.ru/portals.htm. The same site is the place where one can open the portal of the private Dutch media company on -еStart.Ru,
which publishes 18 different periodicals. Among these one can find
Cosmpolitan, Houshold Matters, Drive, Harper’s Bazaar, The Cult of Personality, The Moscow Time, The Sent Petterburg, Vedomdst etc.
In order to be preferred the portal could offer some free services such as e-mail, chat rooms, web space etc. The information to be searched should be aranged in channels under certain subjects. Except of the companies that posses on line services, their own portal services are offerred also by some of the basic directories and search engines. Yahoo began to offer news, AltaVista – on line translation, Northetn Light presents several types business search for information about companies, shares etc.
The vertical market is a special industry or group of enterprises where similar products are researched and developed with use of similar methods of selling. Well known examples in this respects are insurance, real estate, banking etc. These are not only resources for news but they suggest also business services. The vertical portal (Vortal) is an important business tool. This actually is a portal for vertical industry. They serve as a source of news and research, they unite buyers and sellers, submit options for transactions. Most of the portals,includiing the Buglarian one dir.bg use personalised homepage for their users. SageMaker is a vertical portal, i.e. information platform, т.е. on the field of oil and gas delivery, chemical and pharmacy products, services, financial markets. An interesting vertical Russian portal is this devoted to Science. It gives options for selecting hper links to various sites. One of them is - http://www.csrs.ru/Public/Stat/2001/Hed2000.htm, which shows various collections of scientific research. On http://www.aes.org.ru/ - one can see the site of the Association of the European Research which is sponsored by The European Commission. It contains 18 categories.One of the best arranged subject Russian directories is available on http://www.list.ru, which shows a rich list of resources.
On the field of law the first site in the Net is - http://www.law.cornell.edu/lii.html. It is also the first Windows-based Web browser, Cello, The institute of Law suggests more than 7 million findings weekly, more than 40,000 users sessions daily among them 70 foreign countries, more than 180,000 links to LI materials. The site suggests disc based products (disc downloads and CD-ROM ) Collection of Historic Decisions of the U.S. Supreme Court, American Legal Ethics Library , etc.)
Among Bulgarian portals one may mention dir.bg, all.bg, bulgaria.bg, ida.bg etc. IDA.GG is awarded business site

  II.5 Intelligent agents Information retrieval

The intelligent agents are systems which use specialised servers, which Interpret “the behaviour” of the agent and they communicate with other
servers. There are various kinds of these. “Mobile agents” are autonomous intelligent programs, which move through the Net finding and interacting with services “on the user's behalf”. They have genuine navigation autonomy. They could be able to fulfil there performance on every kind of computer. They use mobile code systems as Java and Java virtual engine.
On http://agents.umbc.edu/Applications_and_Software/ one may see hyper links to various sites of various agents. If we select Software an option appears to make a choice among angents for applications for electronic commerce. One of these, ShopBot is a marketig agent, which has an address: http://www.clickthebutton.com/ . It is decisive marketing assistant which works on the background and compares prices for millions of of retailers and hundreds of stores and shops.

On http://webrom.net/search_and_find/ it is possible to select among hpper links to sites of intelligent agents. The user has an option to register himself on this address in order to receive free information for the market status within a period from two weeks for up to 10 companies.
The marketing agent for price comparing PriceComparison.net (www.deatime.com), is capapable to compare the models and prices of various suppliers of various goods, for example Desktop Computers.
According to some researches the marketing through conventional devices is four times more expensive than the marketing through Internet..
The query for “Search Engines Features" OR "Meta* Search Engines" gives 14214 findings. For the same query meta- engine Beaucoup gives 15 sites. (The first results are found by Look Smart и MSN Search). The search is done in parallel by 10 search machines and directories.
Some robots fulfil also the functions of creating and support of the mirror images of sites. When the software archives are for example on many sites this means that more of the users from a certain country or region have faster access and to be able to find and copy for example certain freeware programs.
The challenges before mobile agents are able for browsing on slides of the presentation which gives the expectation for the moving agents, factors, they depend on – unpredictable surrounding, low speed or busy net links, data sources are often doubled by many computers.
For use of other technics and strategies among which is use of Regional and National search engines eventually will be revealed in the next materials.