| <Contents | ||||||||||
|
VI Google Answers http://answers.google.com/answers/ is an answering service that permits users to ask a pool of researchers questions in return for a fee of between $2 and $200. Questioners set the amount of money they are willing to pay, the precise amount depending on the question’s complexity, how quickly an answer is required, and the value of the answer to them. The questioner is the final arbiter who decides whether or not the question has been answered satisfactorily. Standardized response guidelines aim to ensure that Google Answers customers receive information in three areas: (i) researched answer, (ii) helpful links, and (iii) details of search strategy used. A recent Cornell study undertaken by reference librarians compared the university’s free e-mail reference service with Google Answers on the basis of 24 questions (see The Google “answer machine” vs. library reference services). Another interesting article from a librarian’s perspective, by Jessamyn West, a freelance librarian who became a researcher for Google’s online question-answering service, appeared a while back in Searcher (vol. 10, no. 9, October 2002) “Information for Sale: My Experience With Google Answers” http://www.infotoday.com/searcher/oct02/west.htm, in which she describes how the Google Answers service works from the point of view of researchers and how they earn their fees. She was subsequently notified by Google that by publishing an article about Google Answers without first submitting this to Google Answers editors for review she had broken a company policy that required all such articles first to be cleared through Google’s corporate communication department. Her researcher status was subsequently revoked. She writes about this experience, and her exchange of correspondence with Google, in a further piece in Searcher (vol. 10, no. 1, January 2003) “Google Answers Back Or How to Become an Ex-‘Google Answers’ Researcher” http://www.infotoday.com/searcher/jan03/west.shtml.
According to Google, it currently (July 2004) draws on a pool of more than 500 “carefully screened researchers”, who are hired as contractors and are paid on a per-question basis. Google says that, while researchers are not necessarily experts in the field related to the question, they are experts in locating hard to find information on the Web. They also add that Google Answers “is not a substitute for professional advice or counsel” http://answers.google.com/answers/faq.html.
Prior to signing up, researchers must go through an application process that tests their research skills and the quality of their answers. An ability to find information quickly on the Web is a key requirement for becoming a Google Answers researcher. Researchers are asked to comply with a code of practice and research guidelines, https://answers.google.com/answers/researcherguidelines.html, which stipulates, inter alia, that researchers may not endorse, promote or recommend any particular product or service, or any manufacturer, distributor, company or service provider. They must also sign and agree with the terms of conditions outlined in a Contractor Agreement https://answers.google.com/answers/researchercontract.html.
Additionally, Google provides a detailed Researcher Training Manual http://answers.google.com/answers/researchertraining.html that offers style and citation guidelines and sets out the dos and the don’ts that are to be observed by researchers. 75% of the price charged goes to the researcher who successfully answers the question, and the remaining 25% is retained by Google to support the service. People who pose the questions are encouraged to rate the answers they receive by a system of stars, with the maximum 5-star rating indicating a highly satisfactory answer, and a 1-star rating given to a poor answer. Those who ask the questions that have been successfully answered can tip the researcher a further small amount over and above the agreed fee, and the tip is retained in full by the researcher. Google says it monitors the quality of the answers sent by researchers, and those who receive too many negative ratings will be removed and will no longer be allowed to respond to questions as part of the Google Answers service.
Google Answers researchers are completely anonymous, and Google does not differentiate between experts in a particular field or those that are simply expert Web searchers. To protect privacy everyone is identified by a chosen nickname, e.g. zorro-ga, and personal information will not be revealed to either the researcher or the community of registered users at any time.
All questions are publicly viewable on the Google Answers Web site for browsing and searching. The interface also allows for comments by other people, not approved researchers, who can chip in with additional information or suggestions.
When you post a question to Google Answers one of the researchers will search for the information you want. When they find it, they will post it to Google Answers, and you will be notified via e-mail. While both researchers and question-askers must obtain a log-in name, you don't need to establish an account to read other people's questions, answers or comments. Users of the service will be charged for the question only if and when an answer is posted to it. If questioners are not happy with the answer provided they pay only a nominal listing fee. One important characteristic of the service is the fact that questions that researchers are currently answering are “locked” to ensure that only one researcher works on them at a time. At the moment, questions are locked for four hours unless the question price is $100 or more, in which case the question is locked for eight hours. During this period, no other researcher can answer the locked question, nor can other users make comments. Once the researcher responds to the question, its status changes from "locked" to "answered". If the researcher fails to answer the question within the locked time period, the question's status reverts to "unanswered" and another researcher can choose to answer it. During the period when the question is not “locked” the price for the question can be increased. Researchers can seek clarification with question-clarification postings, and question-askers can also ask for clarification of or amplification to answers provided. The Google Answers opening menu offers a link to a log-in page where you create your account, and a box to enter your questions. If you simply want to browse or search there is no need to log in and you can browse previously asked questions or search the entire Google Answers archive. The opening page shows a link to view all questions currently awaiting answers. Clearly the number of questions is increasing all the time: in mid-January 2004 it showed current 1,642 questions, which by early May had grown to over 3,000 (although there are presumably also seasonal peaks). You can search for words that might appear in a question, or browse by broad topics and a range of sub-topics. However, these are very broad indeed, e.g. one sub-group of Arts and Entertainment is Books and Literature, and a sub-group of Relationships and Society is Politics, but without further sub-categories. Therefore searching initially for specific topics is probably a better approach even when you are merely browsing. Searches can be conducted using the same Google advanced search operatorsdescribed in earlier sections (see Advanced Search & Google search operators). Each results page shows the number of current questions and the subject and nature of the question, accompanied by a one-line summary, the name (an anonymous nickname) of the person who asked the question, the date posted (listing the latest questions first, but this can be changed to list the oldest question first), the current status – whether answered, locked, days left for answer, or expired – the number of comments (if any) on the question posted by others, and the price offered for an answer. Alternatively, you can sort by price offered (ascending or descending) by clicking the Price link. On the top of each page there are three “Sponsored Links” (i.e. small ads) related to the question in one form or another – sometimes rather obscurely, it has to be added. Once a question has been answered it shows the nickname of the researcher who provided the answer, and, for some of them, the rating accorded to the answer. When clicked on, it shows the full text of the question, subject and category, nickname of the person who asked the question, the price, posting and expiry date, and each question has its own unique threadview ID number. This is followed by the answer, and also displayed is any correspondence between the question-asker and the researcher if there was any need for clarification. Sometimes there is quite extensive e-mail correspondence between the two parties. Researchers also provide an indication of their search strategy, although this seems to be confined to Web search strategies for the most part. Google Answers on Africa and African studies topicsBrowsing for topics relating to Africa or African studies is rather time-consuming. I have not conducted an exhaustive search of the archive, but there would not appear to be a very large number of questions on African studies topics at present. Quite a substantial number of questions that are Africa-related have to do with business activities, investment and trade in Africa, and requests for market research; there are also questions relating to travel in Africa, cookery recipes from Africa, and questions on some other more general topics. There are many on African American topics and issues. Spending some time and browsing questions and their answers can be quite entertaining. Many of the answers provided would appear to have taken a fair amount of the researcher’s time and seem to be good value for money for the usually very modest fees paid, but in most cases it would require research and/or peer review to evaluate fully the quality of the answers, and to verify them in terms of their accuracy, reliability and currency. Below are some more general observations:
Below is a sampling of a range of fairly recent questions relating to Africa or African studies topics. Click on the Question/Thread ID URLs to see the answers (and comments) for all of them. While I provide some of my own comments on the answers, users of this guide will need to judge for themselves how they rate the answers, and how they might compare with answers sought through library reference services. Subject: History of Asians in Uganda, East Africa Question/Thread ID: http://answers.google.com/answers/threadview?id=29825 Google Answers: provides a short synthesis of the history of Asians in Uganda, but with vague bibliographic citation regarding sources used; a very haphazard listing of book and studies on the subject “culled from links on many of the sites found in the search process”. If the Google researcher had suggested that the questioner consult a number of bibliographic sources in print format, e.g. Cherry Gertzel’s Uganda: An Annotated Bibliography of Source Materials (1991) this would have quickly led to a wide range of books and articles on Asians in Uganda, and the expulsion of Asians by Amin, etc.
Question/Thread ID: http://answers.google.com/answers/threadview?id=64000 Google Answers: $160 is an unusually high price for a question of this nature and clearly the questioner, nelly_bly-ga, must have wanted to obtain this book very badly. The answer here is preceded by a number requests for clarification and one initial suggested answer, which turns out to be the wrong book, before super sleuth leli-ga (see also African-Caribbeans in Scotland 1800s below) chips in and comes up with the right answer: the book (published by Goose Lane in 1992) is called Fadimatu, by Jennifer Mitton, a Canadian author who lived and worked in Nigeria. The researcher provides a synopsis of the book, links to reviews, a link to the author’s Web site and to an online biographical profile of Jennifer Mitton, and even tracked down the fact that second-hand copies were available for sale at http://www.alibris.com. The researcher also reports that although the novel was short-listed for the F. G. Bressani prize, it was apparently criticised for “cultural bias” by Dieter Riemenschneider, a German African literature scholar. A thoughtful and well-researched reply that was given a 5-star rating by the questioner. What is extraordinary here is that the sponsored links on top of this Google Answers page advertise not books, as one might expect, but show an ad “Keep Your Joints Healthy”, plus ads for two chiropractic clinics offering their services. What, one wonders, is the connection here? For once the contextually-targeted AdWords don’t seem to work too well! Subject: Afro-American bookstores Question/Thread ID: http://answers.google.com/answers/threadview?id=68109 Google Answers: this seems good value at first
sight, a 9-page listing of African American bookstores with full address
details, telephone and fax numbers, e-mail addresses for a few of them,
contacts for most, although there are only two bookstores with Web sites. The
researcher states that there are a couple of online directories that list
African American oriented bookstores, “however I found the most up-to-date one
is the one I have listed below”. However, we are not told the criteria for
“most up-to-date” nor does the researcher tell us the source, or whether this
listing is in fact a complete copy-and-paste job from the directory considered
to be the most up-to-date. There are actually quite a number of such online
directories, e.g. the African American Bookstores National Directory at http://www.smallbusinesses.com/BlackBookstores1.htm,
and others include http://www.seemeonline.com/publishers/bookstores.html, Remarkably, there is relatively little overlap in these directories and only a small proportion of bookstores are listed in all of them. One might therefore assume that those listed in all the directories are the most important ones, but neither the Google researcher’s listing nor any of the online directories gives any indication which are the most significant African American/Black bookstores in terms of their size and range of stock and services, although bookstores with a Web site might well reflect an operation of some size. Although this question was answered, no star rating is indicated and if $100 was paid for this it seems a waste of money as the same information could have been freely obtained from publicly accessible Web sites. Subject: African studies – Ethiopia – Amharic Question/Thread ID: http://answers.google.com/answers/threadview?id=91523 Google Answers: no answer, but pinkfreud-ga suggests a couple of Web sites on Ethiopia and draws attention to an article about the use of search engines in Amharic. Subject: African westernisation Question/Thread ID: http://answers.google.com/answers/threadview?id=201393 Google Answers: this question generated quite a lot of clarification postings between the questioner, minxie-ga, and the researcher, digsalot-ga. At one point, ominously, the researchers asks, “do you need the answer in essay form?” A fairly full answer is provided based on information pulled from a number of Web sites, and quoting extensively from Peter Beinart’s Out of Africa, which is duly acknowledged. The second part of the answer focuses mainly on Nigeria and its ethnic groups and contains a lot of quoted matter from an online article, “Ibo Tribe” by Khoi Ta, posted on a Usenet group. The researcher then goes on to state that Chinua Achebe is “the main source for information about Ogidi currently available for scholar and student” and recommends a number of Web sites on Achebe. Not everybody might agree with this, but on the whole this is quite a good answer. Minxie-ga is very satisfied, gives it a 5-five star rating, and adds a $20 tip. Subject: African-Caribbeans in Scotland 1800s Question/Thread ID: http://answers.google.com/answers/threadview?id=226458 Google Answers: there is an interesting exchange of e-mail correspondence here between the questioner, kyraeh-ga, and the Google researcher, leli-ga, apparently resident in Edinburgh. The researcher is clearly intrigued and motivated by the challenge of tracking down information. Part of the answer includes informative profiles of African-Caribbeans who lived in Edinburgh in the late 1880s. This is supported by a wide variety of well-documented sources, including books, periodical articles, and archival as well as online resources. All quoted material is duly acknowledged with proper citation. The researcher also assists the questioner to track down a thesis on the topic and suggests how it might be accessed. Here is a researcher who has not just done a quick search of Google, but has tracked down books, has checked Scottish university and library Web sites, and even the Scotsman newspaper archive. A labour of love, and a true 5-star answer. Subject: Chemistry of the San arrow poison Question/Thread ID: http://answers.google.com/answers/threadview?id=298778 Google Answer: this is really more of a scientific rather than an African studies question. The answer would appear to be well informed, with links to relevant full-text studies (although two of them are subscription-based, requiring payment of a fee). There are also good suggestions for sources as they relate to the symptoms and the action of the toxin. A clarification posting of the answer provides further information, and links, on the chemical properties, the definition of terms, and points out some caveats. The Google search strategies are very clearly set out. The questioner found the answer “generally superb” and gave it a 5-star rating. Subject: Environmental water issues in the country of Burkina Faso in West Africa Question/Thread ID: http://answers.google.com/answers/threadview?id=319013 Google Answers: the Google researcher indicates that he/she cannot find the above quote, but provides links to alternative sources with some helpful explanatory notes how to track down the information. Rated 5-star by the questioner. Subject: Educational free books Question/Thread ID: http://answers.google.com/answers/threadview?id=323552 Google Answers: when questions are poorly phrased it can lead to confusion, as in this case. At first glance this question would seem to indicate that the questioner has surplus books that he/she wishes to donate to Nigerian recipients and seeks assistance with shipping, which results in a request for clarification and a suggestion to consult with the Africana Librarian Council’s Book Donations Committee that assists with shipping to books to select recipients in Africa at (see http://www.albany.edu/~dlafonde/Global/bookdonation.htm). But it later emerges that the questioner, presumably resident in Nigeria, actually wants to receive free books in the field of computers and the Internet and apparently lacks the hard currency to pay for them, although is willing to fork out $40 for an answer. Subject: Search for “a male singer from Senegal.” Question/Thread ID: http://answers.google.com/answers/threadview?id=324596 Google Answers: a good straight-forward answer referring the questioner to the Orchestre Baobab (rather than a singer of such a name, who doesn’t exist), the orchestra’s official Web site, the link to the All Music Guide write-up on the band, and the Amazon.com listing for the group. One couldn’t ask for a great deal more for $2!
Question/Thread ID: http://answers.google.com/answers/threadview?id=328349 Google Answers: an amusing bit of banter (reminiscent of an exchange of views on Nigerian Usenet groups) that also attracted 11 comments from other people chipping in for fun. Rated 5-star, with a $5 tip thrown in, “for handling all those awkward questions with the skill of a very accomplished politician.” However, how is it that the Google Answers editors, who are supposed to monitor all answers, allow this sort of posting?
Subject: Conflict resolution principles and the Truth and Reconciliation Commission Question/Thread ID: http://answers.google.com/answers/threadview?id=337487 Google Answers: questions, like this one, can be very long and demanding yet offer little financial reward. There is a full and well-constructed answer, 13 pages long, with some useful tips and, importantly, including a measure of evaluative comment. There are also helpful clarifications of the answer, in which the researcher draws attention to additional sources. Although rated 4-star, apparently the questioner found the information a little overwhelming. But it certainly looks like very good value for $15. Subject: British invasion of Zanzibar Question/Thread ID: http://answers.google.com/answers/threadview?id=344305 Google Answers: the researcher bases his answers on information found on three Web sites: the first three are from a site hosted by the Zentrale für Unterrichtsmedien im Internet e.V. in Germany and which in turn (for East Africa-related topics) are based on the East Africa pages of the World History at KMLA site, an acronym for the Korean Minjok Leadership Academy, an elite boarding high school located in countryside South Korea. The second two sources cited are from the pages on Tanzania and Zanzibar from the Flags of the World Web site at http://flapspot.net. Hardly the most authoritative sources, but then what can you expect for $3?
Overall, Google Answers is quite an enterprising and useful service, but its main limitation is the fact that, for the most part, “answers” are restricted to information that can be found and referenced in freely accessible online resources. Google researchers may well be able to provide useful suggestions and recommend appropriate Web resources as a starting point for research, but they can rarely offer any in-depth content or analysis. While the above examples show that answers can sometimes be thoughtful and well researched, I would still be hesitant to recommend Google Answers for authoritative and reliable answers to questions as they relate to African studies research, for which it will be much more judicious to consult with the reference staff at African studies libraries – quite apart from the fact that librarians provide their services for free. Indeed, it would be interesting for the Africana Librarians Council (ALC) of the African Studies Association to conduct a similar exercise to that described in the Cornell study mentioned earlier (see The Google “answer machine” vs. library reference services), with test questions relating to African studies research topics. Google Web Alerts http://www.google.com/webalerts is a new service launched in April 2004 and is currently still in its Beta version. It works in much the same way as Google News Alerts except that, instead of alerting you to news stories, it alerts you about topics of interest on Web pages; it can also monitor references to your name, or the name of authors/academics, the title of a book or paper, etc. You sign up on the Google Web Alerts page and the verification/confirmation process works in the same way as Google News Alerts. Like the latter and for general Google Web searching, you can tweak your search terms by using some of the operators described in the previous pages including, and especially, the use of double quotation marks. You can select to receive alerts daily or weekly. Potentially, Google Web Alerts could be more useful for Africanists than the Google News Alerts, because items about specialized African studies topics are more likely to be found on Web pages and Internet documents than to be reported in major international news media. Google promises to send you Google Web Alerts “when there’s new information on the Web matching the search you specify”. However Google’s use of the word “new” is somewhat ambiguous. Tests carried out (in April and May 2004) for a variety of Web alerts have been disappointing. Google states that, as for Google News Alerts, you can incorporate Google Advanced Search techniques to fine tune your Google Web Alert settings, by selecting the conditions you want on the Googleadvanced search pages (see Advanced Search & Google search operators). However, for a number of experiments conducted over a period of several weeks, it submitted alerts of largely irrelevant items, despite my using special search operators. Test searches and alert terms containing words such as “African studies” – despite using “+” or “-” operators or enclosing terms in quotation marks – resulted in alerts to a variety of Web sites or documents that somewhere contained the word “African” or “African American”, nor were they necessarily “new”. For example, in order to be alerted about new reviews or other Web page listings about the African Studies Companion, requesting alerts for either +african +studies +companion or “african studies companion” with quotation marks, initially generated alerts for, among others, the banning of imports of African rats, African droughts, African safaris, or African American men. However, inexplicably, alerts then improved over time and returned far fewer irrelevant results. A Web Alert request for “africana librarians council” (ALC) did rather better and generated three alerts over a period of the two months of April and May 2004, one an alert about the forthcoming Spring 2005 ALC meeting at Northwestern University, another about the Spring 2004 meeting held at the University of Michigan, and the third about the ALC’s Directory of Book Donation Programs. This directory has, of course, been accessible online for some time now, but presumably Google generated an alert as the directory was updated in March 2004. Unfortunately Google then spoilt it by sending a further alert, in July 2004, about the Africana Librarians Council meeting in Bloomington in Spring 2001! For other tests, alerts received related to references and citations that were far from “new”, and had been on the Web for a long time – for example, items relating to listings on some of the major African studies portals and directories, or those of online booksellers and dealers. In one case, it generated alerts about online book reviews that appeared as long ago as 1995, and in another it sent alerts, received in triplicate, about items that had been published on the Web in 2002. Presumably the reason they were interpreted as new was because the Web sites had undergone minor changes and Google robots revisited and re-indexed the pages. So, for the time being at least, this Google service does not have a great deal to recommend it for African studies-related topics and research. Google Personalized Web Search Another new service that made its debut in April 2004 was Google Personalized Web Search http://labs.google.com/personalized, which promises to use personal preferences to deliver search results based on interests selected by the user. In order to establish a personal profile, the Google Personalized home page offers a range of 13 broad categories that, when ticked, lead to slightly more specific sub-categories. For example, “Regions” leads to Africa and ten other regions; “Society” leads to 12 sub-categories including Education, Government, History, Politics, etc.; Arts/Cinema leads to 8 sub-groups, including Architecture, Art History, Performing Arts, etc. You then tick a check box to indicate your categories of interest on the sub-menu. On the basis of the profile established you then start searching. On the first page of the search results you will see a slider at the top of the page, which can be dragged from no personalization to maximum personalization or somewhere in between, and the results displayed change as the degree of personalization changes. Google says it does that by a new type of algorithm that dynamically reorders results by weighting the interests you have entered into your profile.Google determines whether or not the results can be personalized based on your preferences. The personalized results are marked by coloured balls next to the results to indicate a higher degree of relevancy based on your profile. Other results show standard Google results. Google saves your profile information, but it can be edited or deleted at any time. Personalized Web search may be considered an exciting development in search technology and could well be the next big thing in the search community. Google Personalized is still in its Beta version and the company says “we are still ironing out the kinks in this beta test”. It has a long way to go before it can be of significant use to the serious academic researcher, although it may work much better for recreational profile interests, hobbies and sports. It is of little use for academic research because, for a start, the profile categories and sub-categories are far too broad. Secondly, Google seems to assume that you are always interested in the same range of topics you have created in your personal profile. Although the profile can be edited or deleted this could become very tiresome, and you might as well conduct a search in Google Web Search. My own initial experiments in Google Personalized Web Search have been disappointing; I find it all rather tedious and time-consuming. Also in April 2004, on 1 April to be precise, in an apparent move to turn up the heat on rivals Microsoft and Yahoo!, Google announced http://www.google.com/press/pressrel/gmail.html that it was testing a preview release of Gmail, a free Web-based e-mail service with much greater capacity and more novel features than its competitors: with a 1GB storage capacity of up to eight billion bits of information (the equivalent of 500,000 pages of email per user), it was to be launched later in the year. Google said the new service would automatically organize e-mail messages by subject, allow users to search their messages by sender, keywords or topic, and promised that each message would be grouped with all its replies and be organized according to conversational threads. Google also stated that it would provide better anti-spam software than its rivals, turning annoying spam e-mail messages “into the equivalent of canned meat”.
The Google press release was faithfully picked up by a vast number major news media such as the Financial Times, the BBC, CNN and almost 500 other news media, but many message boards were convinced that the announcement was an elaborate April Fool’s Day hoax, especially since it coincided with another announcement by Google, on the same day, that advertised for job openings seeking engineers for a “Google Copernicus Hosting Environment and Experiment in Search Engineering (GCHEESE)” lunar outpost. A day later Google admitted that the advertisement for lunar outpost jobs were indeed a joke, but insisted that they were very serious about Gmail, although Google’s PR department had apparently failed to check their calendar before the launch of the new service on April Fool’s Day.
Perhaps so many people were convinced that Google’s announcement was an April Fool’s Day hoax because it appeared to make outlandish claims, and the press release had a rather cheeky tone. If the Google Gmail service becomes publicly available, with its offering of a full gigabyte of storage, it will certainly put them streets ahead of competing e-mail services from Yahoo! and Microsoft’s Hotmail.
I have not put the Gmail program to the test as at the time of writing (June 2004), and it is available only in early test versions. Initial reactions to the test versions have been mixed, but the new Gmail service will no doubt attract huge numbers of subscribers.
However, there is a price to pay for this “free” service. Google hopes to make money from it by programming its servers to search or scan e-mail messages for keywords, against which they hope to sell related advertising, and that will then appear in the messages. For example – presumably using the same technology as that for the “sponsored links”, the small text advertisements that appear on Google Web Search results – an e-mail to a friend saying you are planning to visit Dakar might result in an ad from a travel agency offering tickets or package deals popping up alongside your message. This is probably harmless enough for offerings of this nature, but it could be a different matter if snippets of information about you could be picked up from messages of a more private nature. Although Google maintains that it will closely guard the content of e-mail messages, and also prevent its own staff from snooping, the new Gmail service is bound to raise serious privacy concerns.
There still seems to be considerable uncertainty and speculation about this new Google offering. While there is a FAQ page about it at http://print.google.com/print/faq.html that sets out how it works, it does not appear on the Google Labs pages (as at July 2004) or on the Google Web Search page, nor would it appear to have been officially launched other than as an experimental program that is still being tested in partnership with a number of publishers that are participating in it, mostly major trade publishers.
In the introductory pages to Google Print (Beta), Google states that its mission is to provide access to all the world's information and make it universally useful and accessible, and adds, “It turns out that not all the world's information is already on the Internet”. Reference librarians will, I think, concur! Google goes on to say that it “has been experimenting with a number of publishers to test their content online”, and it looks as though Google Print has been released to take on Internet bookseller giant Amazon.com and to encroach on their Look Inside the BookÔ, and its extension Search Inside the BookÔ, see http://www.amazon.com/exec/obidos/tg/browse/-/10197041/002-6795202-3108869 and http://www.amazon.com/exec/obidos/tg/browse/-/10197021/002-6795202-3108869 that offers full-text access to view over 33 million pages within 120,000 books at present. However, the new Google service differs from this in as far as Google indexes only fairly short excerpts from the books for the most part. This includes text pulled from cover blurbs, contents pages, introductory and preliminary matter, author biographies, and sometimes chapter extracts, for which Google says it is ranking content using the same page ranking systemused in rating Web sites (see Google’s page ranking and indexing system). Pages that are part of the Google Print program are identified with print.google.com/print at the beginning of their URLs and have an "About Google Print" link at the bottom of each page. These pages are also clearly marked in Google's search results with a [BOOK] tag, similar to those used to identify non-HTML search results such as PDF, etc., and the ISBN number is indicated for each book title. Links are available from each results page, providing users with a quick route to online retail outlets from which they can purchase the book or other content. Google says bookseller links are not, at this time at least, paid for by those sites, nor does Google benefit if users of Google Print make a purchase from one of these retailers.
In March 2004 the Beta Google Print project also added some magazine articles, for which the pages are marked [MAGAZINE], but at the moment the choice is still very limited indeed and most of them would appear to relate to magazines published by the Reed publishing conglomerate.
Searches can be activated (using the standard Google Web Search box) by typing site:print.google.com followed by the search terms – for example, the name of an author.
Example: site:print.google.com wole soyinka produces 22 results from print.google.com, of which 16 are book-related and the remaining six are cover articles on Wole Soyinka in Variety, Publishers Weekly and Library Journal. However, part of the listings relate to books which contain merely references to Wole Soyinka in extracts from critical studies, rather than works by Soyinka himself – for example, a book on the mythic imagination of Africa, or references in publishers’ blurbs, such as that for the Killam and Rowe Companion to African Literatures.
Result number 1 for the above example is Conversations with Wole Soyinka by Biodun Jeyifo (University Press of Mississippi, 2001) and offers four pages of excerpts from this book. The excerpts for most other book titles also have 4 – 8 pages each (except for one, which only has a single-line review), a number of them also including review extracts and biographical notes.
In some cases it is not clear from where precisely in the book the excerpts are quoted, but presumably the publishers provide the excerpts to Google Print. In one case the complete chapter one of Soyinka’s The Open Sore of a Continent: A Personal Narrative of the Nigerian Crisis (1996) is made available.
Whether Google Print will take off in earnest remains to be seen. Google/OCLC - Open WorldCat pilot
The Online Computer Library Center (OCLC) is a (US) non-profit, membership, library computer services and research organization, dedicated to facilitating access to the world’s information and reducing information costs. It links more than 45,000 libraries in the US and in 83 other countries around the world. OCLC services help libraries locate, acquire, catalogue, access and lend library materials. Its massive WorldCat database contains over 55 million bibliographic records, merging catalogues of libraries around the world.
The Open WorldCat pilot http://www.oclc.org/worldcat/pilot/ is a year-long initiative designed to make library resources available from non-library Web sites. The pilot aims to test the effectiveness of Web search engines in guiding users to library-owned materials, making libraries more visible to Web users, and more accessible from the Web sites and search engines to which many people nowadays turn to first. The Open WorldCat uses limited fields from 2 million records, a subset of the more than 55 million records in WorldCat and makes them accessible via Google or Yahoo! http://www.yahoo.com/search. The pilot scheme commenced in December 2003, and in July 2004 the OCLC announced that – based on feedback from pilot participants and focus groups, as well as click-through statistics and other metrics – it deemed the Open WorldCat a success, and that it will continue to extend and refine functionality of the pilot while plans for a full production version are made. Searches on Google (or Yahoo!) retrieve the records and link through OCLC to library holdings. At present the subset targets only the most popular and widely available books, selecting titles held in at least 100 libraries. Approximately 12,000 libraries are participating in the scheme, most of them in North America. To locate WorldCat records on Google or Yahoo! you can use three methods:
OCLC says that current page rankings for showing records are not indicative of the final rankings that will be in place when all records have been properly indexed and formatted by Google and Yahoo!.
In the pilot scheme it is the most popular authors that are most extensively represented, much of it in fiction. For example, a search wole soyinka "find in a library" finds 141 results, but more specialist African studies monographs or reference resources are less widely represented.
For example, for the prolific author Bernth Lindfors, author and editor of numerous monographs, edited collections, and reference works on African literature, a search for bernth lindfors "find in a library" shows library holdings of only two of his titles at present, including his Black African Literature in English, 1992 – 1996, published in 2000.
In a search for african studies companion “find in a library” Open WorldCat finds holdings of both the first and second edition of this title (2nd ed. 1997, 372 holdings), but not the third edition published in August 2003, presumably because this is a fairly recent title.
For each result clicked on – in the Soyinka example above, the first result, as in Google Print, is Conversations with Wole Soyinka by Biodun Jeyifo (University Press of Mississippi, 2001) – WorldCat provides details of author, title, publisher, publication date and ISBN. You then enter a postal code, state, province or country to find libraries holding the item. In the Soyinka example above, entering, e.g. “NY” or “New York” will show 37 public and academic libraries in New York state holding this item in their collections. For the Lindfors bibliography it shows 12 academic libraries in New York state.
For most libraries there is a link to its online catalogue (OPAC), an indication of the city, state/province (or country), a link to a location map, and an indication of the type of library, i.e. academic or public. For titles where it can find holdings in only a few libraries in the state selected, it also shows details of other, nearby regional libraries that hold the book.
In this pilot version at least, you cannot enter a command to request the database to show the holdings of all libraries throughout the United States for any title. However, there is a trick for getting round this: by entering a postal code or state where the item is not likely to be held, it will come up with listings of library holdings elsewhere. For example, entering “DC” for the Lindfors title indicates “No libraries found near you”, but then offers regional library holdings in Maryland and Virginia, plus a listing of holdings in almost 140 libraries in other US states, together with a small number in Europe and elsewhere.
This promises to become an enormously useful service. While access to WorldCat can be had at most academic libraries who subscribe to OCLC services, they are less easily accessible to those outside the traditional library environment. Many students and academics, especially those outside North America, have no means of easy access. This is therefore a most welcome new service, and will be even more so when the full production version becomes available.
While holdings information is still primarily focusing on North American libraries at present, WorldCat is also extremely useful to two other groups in the book chain, the authors and the publishers, who are not normally able to gain access to the WorldCat database to determine which libraries have purchased their books. Google Alert (not part of Google) Somewhat confusingly, Google Alert http://www.googlealert.com is not affiliated with Google, but uses Google's Web services API™ (Application program interface or application programming interface) to perform its searches. Founded in 2003 by Gideon Greenspan, a Ph.D. student in the Laboratory of Computational Biology at Israel's Technion, the Israel Institute of Technology, Google Alert is developed and marketed by Indigo Stream Technologies Ltd., a Gibraltar-based company. It functions very much like Google Web Alerts except that it seems to perform rather better. It runs daily personalized Google searches for you and sends them to you by e-mail when the query produces new results. You can use it to keep track of anything on the Web, including information about yourself (or what they call “ego searching” nowadays), keeping on top of special interests, research projects, new book titles, book reviews, or anything else you wish to monitor, except that it does not cover Google News, Google Groups, or Google Image Search at the moment (July 2004). You create your alert by entering a username, password, and e-mail address, and then enter the searches you would like Google Alert to perform every day (using double quotation marks for searches for individual names or phrases). You will then receive an e-mail alert whenever a new site appears in the top 50 results for your searches. The initial e-mail notification includes up to 50 results per search, which you are likely to have seen before, but Google Alert promises that subsequent alerts “will include only new results that have not been reported before.” You can also view your results online by clicking “Browse results” in an integrated mini-browser. You can chose to receive alerts in HTML format, including clickable links both to the original site and Google’s cache. Other delivery options are by RSS feeds, if you have an RSS reader (RSS, Really Simple Syndication, is a dominant XML format for distributing news headlines on the Web), or via TrackBack (a popular utility that lets your Weblog display Google Alert results and which is supported by blogging systems such as Movable Type, Userland and Blosxcom). The feed settings can be changed at any time, as can the search settings, which offer various options for frequency of searches. You’ll have to remember, however, that Google Alert is dependent on Google for results, and you will receive relevant alerts only when Google updates its index (see also Google’s page ranking and indexing system). You can refine searches using the search settings and the Advanced Search pages, allowing you to include or exclude terms, and to filter by language or domain, etc. The basic (or “Current”) Google Alert service is free and allows you up to three alerts and up to a total of 150 results for all your queries. Charges apply for advanced services, which let you search deeper and more widely, and which offer more tracking power (i.e. number of search terms, top Google results tracked, and daily search capacity) and a host of advanced features, including targeted searches by language/country, personalized relevance, preferred search time, an ability to filter for precise capitalization or punctuation of search terms, additional e-mail recipients, and more. The current (July 2004) charges are, for the second level service (“Personal”), $4.95 a month, $9.95 for “Premium”, and $19.95 for “Professional”. The advanced service introduced in April 2004 also makes use of a new SightPoint personalization technology that automatically rates new search results based on their similarity to results the user has clicked on before. SightPoint uses something called “Bayesian statistics” made popular by spam e-mail filters. The Google Alert service has an attractive, clean interface, and one useful feature (which is part of the free service) is that it displays the Google page ranking (see Google’s page ranking and indexing system) for each result reported next to the page title. It seems to have the edge on Google Web Alerts, even for its basic free service, delivering more relevant results and, by using an optimized matching algorithm, can determine whether a search result has been seen before, thus minimizing the chances of seeing repeats. Google Alert says its terms and conditions of use are subject to Google's API™ terms http://www.google.com/apis/api_terms.html, which state that the Google Web APIs service is made available for personal, non-commercial use only (at home or at work). Although I can’t quite see how this can be regulated or policed, I shall therefore recommend it for non-commercial use only! Google Alert is a clever and useful tool.
|
|
|||||||||
|
|
||||||||||
|
|
||||||||||
| ||||||||||
