The ADS Abstract Service is only seven years old, but it is already an indispensable part of the astronomical research community. We get regular feedback from our users and we implement any reasonable suggestions. In this section we mention some of the plans for improvements that we are currently working on to provide even more functionality to our users.
One important part of the ADS Digital Library will be the historical literature from the 19th century and earlier. This part of the early literature is especially suited for being on-line in a central digital library. It is not available in many libraries, and if available is often in dangerously deteriorating condition. The access statistics of the ADS show that even old journal articles are accessed regularly (see OVERVIEW). The ADS is working on scanning this historical literature through two approaches: Scanning the journals, and scanning the observatory literature.
1. We are in the process of scanning the historical journal literature. We already have most of the larger journals scanned completely. Table 11 shows how much we have scanned of each of the journals and conference proceedings series for which we have permission. The oldest journal we have scanned completely is the Astronomical Journal (Vol. 1, 1849). We plan to have the Monthly Notices of the Royal Astronomical Society on-line completely by early in 2000. After that we plan to scan the oldest astronomical journal, Astronomische Nachrichten (Vol. 1, 1821), Icarus, Solar Physics, Zeitschrift für Astrophysik, Bulletin of the American Astronomical Society, and L'Observateur, and the conference series of IAU Symposia. Other journals that we plan to scan are the other precursor journals for Astronomy & Astrophysics, if we can obtain permission to do so. We will also scan individual conference proceedings for which we can obtain permission.
| Journal | Scanned Volumes | Publication Dates |
| Acta Astronomica | 42-48 | 1992-1998 |
| Annual Reviews of Astronomy and Astrophysics | 1-33 | 1962-1995 |
| Annual Reviews of Earth and Planetary Sciences | 1-23 | 1973-1995 |
| Astronomical Journal | 1-114 | 1849-1997 |
| Astronomical Society of the Pacific Conference Series | 1-5, 7-22, 24-63, 65-69 | 1988-1994 |
| Astronomy and Astrophysics | 1-316 | 1969-1996 |
| Astronomy and Astrophysics Supplement Series | 1-120 | 1969-1996 |
| Astrophysical Journal | 1-473 | 1895-1996 |
| Astrophysical Journal Letters | 148-473 | 1967-1996 |
| Astrophysical Journal Supplement Series | 1-107 | 1954-12/1996 |
| Baltic Astronomy | 1-5 | 1992-1996 |
| Bulletin Astronomique de Belgrade | 153-155 | 1996-1997 |
| Bulletin of the Astronomical Institute of Czechoslovakia | 5-6, 9-42 | 1954-1955, 1958-1991 |
| Bulletin of the Astronomical Society of India | 8-23 | 1980-1995 |
| Journal of the Korean Astronomical Society | 1-29 | 1968-1996 |
| Journal of the British Astronomical Association | 92-107 | 1981-1997 |
| Meteoritics and Planetary Science | 1-33 | 1953-1998 |
| Monthly Notices of the Royal Astronomical Society | 1, 100-301 | 1827, 1950-12/1998 |
| AAS Photo Bulletin | 1-43 | 1969-1986 |
| Publications of the Astronomical Society of Australia | 1, 3-13 | 1967, 1976-1996 |
| Publications of the Astronomical Society of Japan | 1-50 | 1949-1998 |
| Publications of the Astronomical Society of the Pacific | 1-109 | 1889-1997 |
| Reviews in Modern Astronomy | 1-9 | 1988-1996 |
| Revista Mexicana de Astronomia y Astrofisica | 1-32 | 1974-1996 |
| Revista Mexicana de Astronomia y Astrofisica Ser. de Conf. | 1-6 | 1995-1997 |
| Skalnate Pleso, Contributions | 11-14, 16-19, 21-28 | 1983-1986, 1987-1990, 1991-1998 |
2. One very important part of the astronomical literature in the 19th century and earlier were the observatory publications. Much of the astronomical research before the 20th century was published in such reports. We are currently collaborating with the Harvard library in a project to make this part of the astronomical literature available through the ADS. The Harvard library has a grant from the National Endowment of the Humanities to make preservation microfilms of this (and other) literature. This project is generating an extra copy of each microfilm. We will scan these microfilms and produce electronic images of all the microfilmed volumes. The resolution of the microfilm and the scanning process is approximately equivalent to our 600 dpi scans. This project, once completed, will provide access to all astronomical observatory literature that is available at the Harvard library and the library of the Smithsonian Astrophysical Observatory from the 19th century back. For the more recent observatory literature we will have to resolve copyright issues before we can put it on-line.
In order to complete our data holdings, we still need issues of several journals for scanning. A list of journals and volumes that we need is on-line at:
http://adsabs.harvard.edu/pubs/
missing_journals.html
In order to feed the pages through a sheet feeder, we need to cut the back of the volumes to be scanned. If they have not been bound before, they can be bound after the scanning. If they have been bound before, there is not enough margin left to bind them again. If you have any of the journals/volumes that we need, and you are willing to donate them to us, please contact the first author of this article. We would like to have even single volumes of any of the missing journals.
The capabilities of the search system and user interface have been developed in close cooperation with our users. We always welcome suggestions for improvements and usually implement reasonable suggestions very quickly (within days or a few weeks). Because of this rapid implementation of new features we have hardly any backlog of improvements that we want to implement. There are currently two larger projects that we are investigating.
1. We plan to convert all our scanned articles into electronic text through Optical Character Recognition (OCR). In order to be able to search full text articles we will need to develop new search algorithms. Our current search system depends on the abstracts being of fairly uniform length. This caused some problems at one time when we included sets of data with abstracts that were 4 - 5 times as long as our regular abstracts. These long abstracts would be found disproportionally often in searches with many query words (for instance in query feedback searches), since they generally matched more words. We had to reduce the sizes of these abstracts in order to make the searching work consistently with these data sets. OCR'd full text will require new search algorithms to make the search work at all. We currently estimate that the implementation of the full text search capability will take at least 2 years.
2. The scanned articles frequently contain plots of data. For most of these plots the numerical data are not available. We plan to develop an interface that will allow our users to select a plot, display it at high resolution, and digitize the points in the plot by clicking on them. This will allow our users to convert plots into data tables with as much precision as is available on the printed pages. This feature has been made available after this article was accepted for publication.
We are currently investigating several new data compression schemes that would considerably reduce the size of our scanned articles. This could considerably improve the utility of the ADS Article Service, especially on slow links. We plan to be quite conservative in our approach to this, since we do not want to be locked into proprietary compression algorithms that could be expensive to utilize. At this point we do not have any time frame in which this might be accomplished.
Copyright The European Southern Observatory (ESO)