FYIFrance: networking and Internet at the BibMun de Lyon (pt.2/2) (continued: by Christian Ducharme -- translated by Jack Kessler) In March, 1993, the public had access to a dozen CD-ROM titles. Usage statistics demonstrated that the public responded well to the service. In the course of the last nine months of the year, there were more than 12,500 consultations at the four posts scattered throughout the library. The average consultation time of a CD-ROM was 14 minutes. In 1994, the number of titles doubled; see the list accompanied by the statistics. List: use statistics of CDROM -- 1994 [I thought non - French readers might be interested to see some of these titles. jk.] Title Consultations Title Consultations AFP 1066 Instruments de musique 420 BGF / Bibliog.Ge'n. Fr. 1893 Kompass France 7553 CD Atlas de la France 284 Kompass Europe 599 CD Actualite' 5303 Lexilaser 1158 CD Litte'rature 728 Maps and Facts 1245 CD The`se 1071 Megastat (France) 578 Code civil 662 Le Monde 3204 Discotext 421 Myriade 2742 Electre 1251 Robert (dictionnaire) 945 Film index 666 Megastat (Europe) 331 Francis 575 _____ Harrap's 228 Histoire au jour le jour 1216 Total 36367 We pay particular attention to statistics. In the policies governing the CD-ROM server, it was firmly set out that the software should be able to furnish precise statistics. In a library, these are important management tools. User statistics are used for the development of the collection and for determining the distribution of workstations in the different rooms of the library. Strict rules as to their collection were established. For example, a user had to pursue a consultation for longer than sixty seconds before that statistic would be taken into account. In this we were thinking of the user who simply couldn't use the CD-ROM (committing errors with the mouse, for example) and who therefore abandoned a search. >From the beginning of this year, the following titles were added: Anne'e strate'gique, Bacchus, Cine'mania, Jour J, Me'moire et actualite' en Rho^ne - Alpes, Music index, Normandie, Pascal et Repe`re. The installation now thus includes 33 titles. >From the first, the librarians took charge of this CD-ROM consultation. The library's information service only provided beginning instruction: the explanation of different search software packages, the basic operation of the medium, setting parameters for printing, etc. The librarians of each department later deepened their knowledge of CD-ROMs by creating small workshops in which they analyzed the products. They rapidly took on this mode of information searching and have no problem in instructing the public. Digitized Text Documents The digitization of documents can respond to several needs. One can digitize to archive the documents (economies of shelving space), to protect the collections (conservation of rare or precious documents), improve access to certain documents (improved searching distributed to several points on the network). After digitization, it is possible to impose a certain number of improvements on a document. For example, one can delete parts of the document which are not pertinent, tag and rearrange the text, and change the size of the document so that it will occupy a single page on the screen or in printing. In these cases, the document is stored as an image (a file in graphic format). It also is possible to use OCR / Optical Character Recognition. The image file thus is converted into a text file (ASCII or ANSI). The advantage of OCR is double: economies of disk space, and the possibility of searching the text itself (fulltext searching). At the Bibliothe`que Municipale de Lyon, the press files for the Department of Documentation of the Rho^ne - Alpes Region were given priority for digitization. For this service, OCR was not used, this type of treatment not being adequate for newspaper articles. The error rate due to page formatting (titles in large characters, the use of wide characters, columnar text, etc.) did not permit good quality recognition. Otherwise the capture of documents in image files gave the advantage of conserving the page formatting. Thus the document is displayed as is on the screen, and its printing produces a copy which conforms to the original. On the other hand, each news article must be indexed to be retrieved, and the files take up much disk space.But OCR software improves, and SGML is making its way toward becoming a standard. During the next several years, OCR and page format encoding will be included in all electronic document software. It then will be interesting to convert this database to fulltext, to permit fulltext search and retrieval. For twenty years the library has collected documentation on its Rho^ne - Alpes region. Until now, the documents have been photocopied and stored in files, divided among 180 subjects and arranged by years. The user who searched for information had to apply mental gymnastics to discover in which file the desired information might be found. And, in many cases, the user had to consult several files to complete a search. In addition, for security reasons, current - year files were not in open access. The public had to present identification in order to obtain these. The digitization of press files did not economize on time or on personnel. The amount of work seems to remain the same, but it is distributed differently: there are fewer manual tasks and more scientific tasks, like the creation of a thesaurus. There has been, on the other hand, an improvement in the productivity and the security of the files. The gain in productivity shows because for the same effort one can obtain a better use of the documentary archive. A search for a file now may be done by the words of its title, by author, by periodical title, by date and by subject. As for security, the risks of theft and of damage to the files are nil. In the interests of the public we have decided not to index the documents using the RAMEAU system. The public can only understand this documentation language with difficulty. Its mode of indexing is useful only if the search is done by a librarian and not directly by the user. A thesaurus for regional documentation was assembled, strongly inspired by the model of RAMEAU. The experiment was interesting, and the thesaurus, which today includes more than 12,000 descriptors, is well adapted to the needs of the general public. The informatization of the Rho^ne Alpes documentation was a success not only for the public but also for the staff. The librarians have reacted well to the organizational changes. Tasks are different and more exacting from a professional point of view, and therefore of greater value. Digitization of Iconographic Collections The Bibliothe`que Municipale de Lyon possesses important iconographic collections. The Rare Items collection alone contains thousands of interesting images. The Arts Department also possesses prints, posters, photographs... so many documents that the public does not often have occasion to see them except during expositions organized by the library. Digitization is practically the only means of restoring this iconographic heritage to the public. Image documents pose a double problem: on the one hand they are not well cataloged, and on the other hand they are not freely - accessible as they often are rare or precious documents. But interest in images is growing. For example the IRHT / Institut de Recherche et d'Histoire des Textes / Institute for the Research and History of Texts just is concluding a campaign to photograph 12,000 illuminated manuscripts. These documents, which are available as diapositives, are sent to a Kodak laboratory to be digitized. Kodak reruns the image files on CD - PHOTO. From this they are imported into the image database. We are assembling an image - bank, which includes a representative selection of each type of document, from illuminations to modern posters, and including prints of the 16th century and photographs of the 19th century. The treatment of images appears similar to that of press files. But there are two possibilities: either to digitize the document in place using a digital photo apparatus (Kodak DCS 200), or to go through a laboratory intermediary to make a CD-PHOTO. The principal steps: the choice of documents to digitize is made by the Arts Department; the images selected are arranged by format or document type (groups of 100 for a CD-PHOTO and of 50 for digitization in place -- the DCS 200 has a memory capacity of 50 images before downloading); the photograph is marked with the date, the classification number, and the fields corresponding to the inventory (these fields indicate the source of the document); then the librarian fills out an entry, the subject - indexing of which represents the principal cataloging element. We have created a thesaurus for the library for the database of Rho^ne - Alpes Documentation, inspired by RAMEAU and the _The'saurus Iconographique_ of the abbe' Garnier. Internet and the "Virtual Library" The third reason which governed the development of an important network at the Bibliothe`que Municipale de Lyon was support of the global project of virtual libraries. This project, begun in North America, is in the process of overturning the traditional function of libraries. It soon no longer will be possible, if one wishes to respond to the needs of users, to stay out on the fringe of the great network of information exchange like the Internet. The importance of the library and the richness of its collections justify this effort, to make cultural patrimony accessible on the national and international networks. Already, the catalog of the Bibliothe`que de Lyon may be consulted on the Internet (via telnet to bm.univ-lyon1.fr). Within a short time, digitized collections will be accessible using one of the client software packages of the WorldWideWeb / W3 variety. We are working on the installation of a "Mosaic" server, which will permit the creation of hypermedia databases. Navigation on the Internet calls for a particular type of training for library professionals. They must learn to use the services offered on the networks: email and bulletin boards, forums (newsgroups), telnet, WAIS, Gopher, Web. We remain convinced that librarians must learn information searching on the Internet quickly. It is important that they learn to exercise their professional skills in this new environment. For all this, their role will not change. They always will be involved in the analysis of content: only the means of gaining access to that content has changed. Their role in the training or education of the public will become important as the networks grow. The public will need assistance to navigate in this environment, which is not yet formalized and which yet contains very few reference points. In two years of operation, the multimedia network has evolved: initially seen as including six stations, it today counts more than thirty, of which twelve are dedicated to public access. Optical - disk electronic publishing is flourishing and the number of CD-ROM titles possessed by the library increases from month to month. We are beginning to digitize other types of text documents, such as tables of contents, brochures, statistical tables, etc. The database of Rho^ne - Alpes press files will diversify, and the Arts Department of the library will begin the digitization of files on the artists for which documents will come from several different museums. During the course of the year 1995, efforts also will be made to expand the images databank. We soon will make this database accessible to the public at la Part - Dieu [main branch of the BMLyon. jk.], and eventually will offer access via Internet. We also will try to put at the disposition of the public an Internet research tool which will provide access to the catalogs of the principal libraries of the entire world. A long effort remains to be made in the creation of more appropriate documentation tools which will permit the public to search for information on the Internet's networks. This work will be done in collaboration with the librarians, who must select servers responding to the needs of the library. One of the difficulties of the Internet networks is that information searching is not static but dynamic. Information in electronic form evolves rapidly. We never are certain to have the power to recall the same information from the same place. There also are problems of the availability of the information servers: a server always can become overloaded, out of order, or disconnected. How to explain this to the public? The evolution of technology encourages us to give to libraries instruments for the search and retrieval of information which always are more effective in responding to the needs of the public. But putting these instruments into place poses more than simply technical problems. We are concerned here with a transformation in the basic function of the library, and the work habits of librarians, which even causes us to reflect on the survival of libraries in a context of the decentralization of information, on the training of librarians, and on the problems of the rights of authors. January, 1995. *** FYIFrance e - newsletter ISSN 1071 - 5916 * | FYIFrance is a monthly electronic newsletter, | published since 1992 as a small - scale, personal, | experiment, in the creation of large - scale | "information overload", by Jack Kessler. Any material / \ written by me which appears in FYIFrance may be copied ----- and used by anyone for any good purpose, so long as, // \\ a) they give me credit and show my e - mail address, --------- and, b) it isn't going to make them money: if it is // \\ going to make them money, they must get my permission in advance, and share some of the money which they get with me. Use of material written by others requires their permission. FYIFrance may be found via gopher to infolib.berkeley.edu 72 and gopher.well.sf.ca.us , and it is in various online archives: the easiest to use is the PACS-L archive, reached via telnet to a.cni.org , login brsuser . Suggestions, reactions, criticisms, praise, and poison - pen letters all will be gratefully received at kessler@well.sf.ca.us . *** end .