Title: Librarians' Index to the Internet
URL: http://www.lii.org
Publisher: InFoPeople Project
Price: Free
Reviewed: February 22-25, 2001

Content and Design

The tagline for Librarians' Index to the Internet is "By Librarians for Everyone," and it certainly makes me proud to belong to that occupational group. The directory, a brainchild of Carole Leita of the Berkeley Public Library, is the work of 80 competent librarians who donate their time to building and maintaining this exquisite directory of Web resources. I am pleased to see that the members of the indexing and abstracting team are identified and given credit and that the directory entries carry their initials. This is an unusually long review for a Web directory, but it deserves to have homage paid to it and to complement the less technically oriented reviews about LII.

There are thousands of subject guides and directories of Web resources. Those who have not yet created one are likely to be working on one. LII stands out from the crowd thanks to its professionalism, objectivity, reasonable selection policy, informative annotations, excellent organization, awesome interface and search features. Unfortunately, far inferior directories divert traffic from LII, like the absolutely not Awesome Library and Encyberpedia that go laughing to the bank to cash in the advertisement fees of "sponsors."

LII has information for 8,000 free Web resources — a relatively small subset even of the worthy free sites on the Web. However, almost all of the selected resources are the best in their categories. LII is like the non-pretentious gourmet trattoria that offers some exquisite food in a very relaxed, squeaky clean and sunlit ambience, where the staff advises you about the cuisine without either acting haughtily, as a maitre d' in New York, or pretending to be your best friend tossing around French culinary terms, as the pseudo-Tahitian restaurant in Honolulu does.

Scope

It is a daunting question — what to include in a directory out of nearly a billion Web sites? At least for LII, some of the limits were given. As it grew out of a personal project into a tool for the Berkeley Public Library, then luckily became funded by the U. S. Institute of Museum and Library Services, then the Library of California, the primary scope of LII is understandably Californian. Its coverage of California is comprehensive in terms of the most worthy resources. LII has a very good page about the selection criteria and I particularly like the introductory paragraph that explains that "LII Indexers compare the relative value of the resource to the range of information resources on and off the Web."

There are many librarians who can recite the same old criteria along with the clichés about the quality of Web sites, but LII practices what it preaches and really chooses the worthiest sites. In my testing I found only a very few sites — like the extremely outdated Atlapedia and the not so "Awesome Library" — that appear in way too many "selective" Web guides and should have not been blessed by being listed in LII for reasons that I discuss in my accompanying review ("Awesome Library", Péter's Digital Reference Shelf, /peter, March 2001).

Of course, California and Californians are not known to be extremely introverted, so there are many sites included that are not specifically related or limited to the state, but rather represent a topic, genre or person of wide interest, such as soccer, classical music and Johann Sebastian Bach. In these cases, the filtering has to be very strict in order to keep this a manageable project and save the indexers from burnout. With that said, however, I wonder why Bulgaria deserves an entry but not Guatemala in LII? I would venture that there are many more Guatemalans than Bulgarians in California, as well as potential visitors to the heartland of the Mayan culture than to the Balkan peninsula. And yes, there are excellent Web sites about Guatemala.

Categories, Subcategories and Sections

The sites are classified into 40 major categories ranging from Arts to Health to Travel. The major categories are then subdivided into a varying number of subcategories. The Law topics subcategory, for example has more than 50 subcategories, from Affirmative Action to Women's Rights. The Food category has only 15 subcategories, although with a cross reference to Specific Foods and Beverages with nearly 30 of its own subcategories. An exception to the rule, the Seniors category is not further subdivided.

The subcategories are listed alphabetically and, if it belongs to more than one category, it is listed under both: for example, Museums are listed under the Arts and the Travel categories. The bold face subcategory entry under Travel indicates that the entries appear under another category, but this is totally transparent to the user. It is more elegant than the similar function in Yahoo! that identifies such a situation by using the @ sign in front of the name of the subcategory, i.e. @Museums. LII's system gives flexibility to the indexers who can list subcategories under several main ones with an implicit, behind-the-scenes link. For example, Museums is also an entry under the Recreation main category, with an implicit link to the Arts main category. Some of the categories are especially well-done, such as the one on the Disabled, although the category name Disabilities would be better. Some of the categories and subcategories may be arguable, as is the case with any classification scheme. I certainly would not assign a subcategory for Chewing Gum under the Specific Foods and Beverages categories, but I certainly would have a subcategory for Steaks.

The sources are typically grouped into three sections: Directories, Databases and Specific Resources. In some categories or subcategories there is also a "Best of . . ." section. For example, the Lodgings subcategory within the Travel category lists five sites under Directories, three sites under Databases, and 10 sites under Specific Resources. Again, it may be argued why the Travel Exchange site is listed under Specific Resources when it is also a good, browsable directory and a searchable database as part of the Vacation Spot site as LII correctly notes.

This is not a big issue because LII is so eminently searchable, that it is not worth agonizing over the inherent limitations of the pigeonholing aspect of otherwise very useful classification systems, not to mention the difficulties posed by the ever changing features of the sites. A plain HTML site with collections of mile long pages for each letter of the alphabet can change into a keyword searchable site overnight when the navigation by the first letter of the alphabet becomes cumbersome.

Subject Headings

LII assigns one or more subject headings to every resource. The subject headings are from the Library of Congress Subject Headings (LCSH) list, and if necessary they are modified to reflect the actual lingo. For example, LII rightly uses the term "e-mail" instead of the more formal and less common "electronic mail." For good measure, there is a cross reference from the term, "electronic mail," to the preferred term. In addition to "see" references there are "see also" references as well that steer the users to other related terms, such as the entry "Frontier and pioneer life (7) – see also Pioneers," or to a category such as the entry "Environment (30) – see also Environment Topics Page for specific topics." There are about 230 cross references among the 4,400 terms in the controlled vocabulary, somewhat less than half of it is a "see also" type.

All these aids make finding the appropriate terms very easy. The only case when I stumbled was when I first did not find Salvador on the list of subject headings even though it is there under El Salvador without a cross reference. I know that it is one of the few geographic names (like The Gambia, The Hague) that use the definite article as part of their name, and the literary warrant is definitely on LII's side. Still, as a compromise, a cross reference would be justified as I may not be the only one who forgets the article because it is not used in the adjectival format, like Salvadorian army, Salvadorian junta, Salvadorian refugees. But all this is just irresistible smart-alecking in light of the very good subject heading list that I enjoyed grazing; I even downloaded it for a little analysis.

More about Subject Headings

The extent of use of subdivisions is a more important issue because its effect is significant in the browsing phase. I feel that there are too many entries that don't justify subdividing, simply because the main heading itself has only less than a half dozen postings. It would be simpler for the user to make one click and see the 3 or 4 sites instead of clicking on two or three subdivided entries, each of them listing only a single site. While it is very reasonable to have subdivisions (even unconventional ones) for certain terms with many postings, such as the term Searching, perhaps additional subdivisions could be created to thin the overcrowded entry on "Searching – Other search engines and indexes" that has 70 items.

The only subdivision with more postings is "Teaching – Aids and devices" with 74, and this is very telling about what an excellent source LII is for students and teachers. It is followed by the terms "Homework" and "Biographies" with 67 and 64 posting, then "United States – History" (56), "Artist," "Health," "Medicine," "Literature,""Maps," "Rock Music," "Travel," "News," "Children's Literature," and "Consumers " with posting between 40 and 50. The only surprise in this cluster is "Astronomy " with 40 postings and a "see also" reference to the "Astronomy" category for specific topics within it. I would have never guessed that it is such a hot topic.

The next stratum with postings between 30 and 40 items includes "Musicians," "Companies," "Music," "Recipes," "Stocks," "Women," "Investments," "Statistics," "Books – Reviews," "Software," "Kids," "Jazz" (by itself scoring 32), "Law," "Art," "Geography" and "Environment." This perfectly illustrates the profile of a public library. "Weather" belongs to the same league, but for a foreigner like myself, the popularity of weather specialists and reports and their elaborate Las Vegas show-room presentation on TV is still baffling after 12 years of residency, but obviously there must be an existing need for knowing all the gory or glorious details that reel off the tongue of the weather persons as bid amounts at an auction site.

The point is that if anything can be done better in this directory, I think it is the revision of the policy of assigning subdivisions to subject headings. There are quite a number of subject headings that beg for subdivision, such as "Homework – Chemistry" instead of just "Homework." On the other hand there are hundreds of subject heading — subdivision combinations that are singletons, assigned only to one resource. These could be rolled into the subject heading that itself alone is assigned to only a few resources. Because of the ease of searching for subject words (instead of browsing them), this is not a critical issue, but certainly is worth considering.

The annotations

The annotations in LII are always informative, often exceptionally so. For example, in the entry about a surfing site, the annotations include a healthy dose of criticism and often alert the user to a deficiency, such as the excessively long HTML file in a library science dictionary, or a cumbersome feature in an otherwise good product or service, like the handling of imported bookmarks in My URLs. The annotations also mention the presence of a particularly valued feature, such as the single word note "searchable" in the description of the electronic journal, Information Research.

The length of the annotations vary from a few words up to 100 words and more. It is defined by the target site, as it should be. Ten words is enough to describe the El Dorado Public Library's news about Web sites, while Yahoo! Internet Life — that also reviews new Web sites — rightly deserves nearly 100 words.

It is also a good idea to have the initials of the librarian who contributed the record hot-linked; this lets you easily find other records created by the same indexer. There is a roster identifying each of the indexers, their library, and — with a few exception — also their e-mail address, so that you can send them a thank-you note.

The only missing feature that I would like to see in the entries is the date for when it was added to or updated in LII. Sites change very often and may introduce new features that may in turn determine if the user chooses to go to the site. Some entries are dated in the annotation, as is the case for the entry about a very useful travel site that most travel guides don't tell you about, but that you had better know. Unfortunately, according to a note added at the end of January, the site is no longer going to be updated.

The Software

Sometimes you keep going back to a restaurant, even if the ambience is less than perfect, because of the good food and pleasant service. In LII, the content alone would justify returning often. The masterfully implemented user interface is another incentive. It provides awesome ambience and decor through the exceptionally tasteful design, beautiful layout and breezy navigation. You must use it to appreciate it.

The search software is based on the enhanced version of the SWISH software. I have seen a dozen of sites that use this software, but none of them can hold a candle to LII in terms of power and intuitiveness. As I understand it, credit is due for the software development and the software support to the Berkeley Digital Library and especially Roy Tennant, who has been instrumental in its development; also, for the impressive enhancements of the search software, Bill Moseley deserves credit. It was definitely worth the effort. The people at the Berkeley campus of UC has been offering services and resources (Current Cite, LJ Digital) that I use in my Digital Librarianship course as best examples. The LII software is yet another feather in their cap.

You can traverse the canopy of the categories and subcategories with the ease of Tarzan jumping from branch to branch. If you don't find your term immediately in the hierarchy, or you don't know where to enter the rainforest of Web sites, you may display the list of subject headings and browse it alphabetically, jumping from sections A to Z, with a short stopover at section M.

Simple and Advanced Search

If you still don't find the most appropriate category, there are many search options listed under the simple query cell. You can formulate a query using Boolean AND, OR and NOT operators, parentheses to group terms, such as (library or information) and science and research. You may use double quotes for phrase searching, such as "information science research". Stemming of search terms is automatic, but you can turn it off and use the explicit truncation symbol to control stemming for one or more terms only.

The keyword search will find records where your search terms have a match in the title, subject or description field. The subject search will retrieve records that have a descriptor or part of a compound descriptor that match your query term. If you are interested in "surfing" you may be tempted to use this word for a subject search, but you will get many unwanted records about "Internet surfing." In such ambiguous cases, it is better to climb down the category branches from Recreation to Sports, and pick the term Surfing that will return 5 hits related to the sport of surfing. As you see, clicking on the term will create the query: surfing not internet, and the next time you can directly do a subject search with the Boolean operator.

You may limit term matching also to the title, the description or the linking URLs. You don't need to worry about the punctuation inside a long URL as each component of a URL is a separate word in the index and the default operator between your search terms is the Boolean AND. This means that if you remember part of the URL, it will suffice to type those words that you are sure of. So, if you recall a site about world war news, it is enough to enter those three words to retrieve the BBC site, and you don't need to worry about the less-than-mnemonic URL http://news.bbc.co.uk/hi/english/special_report/1998/10/98/world_war_i/newsid_197000/197437.stm

The choices listed above are mutually exclusive, i.e. you cannot limit your search terms to the subject and the URL fields. For this purpose you need to use the advanced query form. Right next to the query cell there is a crib sheet that gives all the necessary hints and shows an example. If that is not enough, there is a longer, very well structured help file with explanations and examples. More importantly, if your query does not retrieve anything, the software automatically displays a friendly page that gives you advise just as a smart and helpful librarian would.

Such features definitely enhance your well-being while using this site. The advanced mode lets you zoom in on items most likely to be relevant for your search. For example, to find all the sites that offer searchable, full-text databases, you create the query: ~de:searchable and ~su:(full-text). This will retrieve 32 items. They are not only perfectly relevant, but also show very informative annotation.

For dessert there is yet another superb page, that offers the best search engines, metasearch engines and directories if you want to follow up your LII search. It could only be better if it were itself a metasearch page where users need only enter the query only and then run it against a group of chosen search engines and directories. Hopefully, there will be another funding opportunity to keep enhancing this superb resource for the benefit of users inside and outside of California.