John Wiley, 2009, -228 p.
Since the early 1990s, Chemical Abstracts Service (CAS) has been developing an intuitive ‘point and click’ interface to scientific information for scientists. Subsequently, SciFinder was released in 1995 as desktop software for commercial organizations, and the academic version SciFinder Scholar became available in 1998. Today these are merged into one product, SciFinder, although there are two interfaces: a client version, which requires the software to be downloaded, and a web version. This book focuses on the web version and on how it works, and suggests how scientists may obtain maximum value from the interface.
This book builds on the text ‘Information Retrieval: SciFinder and SciFinder Scholar’, which was published in 2002. Feedback from that text was positive, but a general comment was ‘I don’t want to read 240 p.’. I understand. Actually I requested the publishers to market that text as ‘about 100 p., with additional figures and diagrams’. The publishers replied, ‘You are the first author ever to ask that a book be marketed with fewer pages that it actually has!’
The biggest changes in the information world since 2002 have been the dramatic increase in availability of full text documents in electronic format and the increased search capabilities on web interfaces. Other changes relevant to scientists have been the introduction of Reaxys and Scopus, and enhancements to Web of Knowledge. Having studied these options, I still am of the opinion that SciFinder offers the best package – provided that users understand how to use it effectively. I hope this books helps.
While it is understandable that scientists really want an interface that requires minimal learning, the reality is that there are numerous issues relating to retrieval of scientific information. Issues start with the content (‘What really is the data behind this interface?’), then go through the difference between author text and indexing (‘Actually what is indexing and how do I learn it’), to the opportunities the search engine provides both at the search and then at the post-processing level. It’s complex, but SciFinder offers simple solutions.
SciFinder is very different from all other information retrieval tools. The most user-friendly alternative search tools, including web search engines, still require some knowledge of truncation, of proximity searching, of synonyms, and so forth. They interpret the question literally and rarely do they offer any guidance on how to proceed. Depending on what the scientist really wanted, they may or may not produce comprehensive and/or precise answers.
SciFinder also works differently in that it guides the searcher. Indeed, when a question is initially asked, SciFinder does not give a straight answer! Instead, it guides the searcher by producing a set of options. SciFinder tells the user to ‘go down this option and you will find a specified number of records, whereas this other option provides a different number of records’. In other words, the user chooses a path based on the actual number of hits, but the choice is not irrevocable and the user may always return to narrower or broader answer sets.
Once a path has been chosen, SciFinder has creative analyze options. For example, for initial bibliographic answers SciFinder provides histograms of the different index terms, or document types, or author names, or publication years. Armed with a lot of information up-front the scientist chooses refinement paths accordingly. In chemical structure searching, SciFinder automatically interprets the query to allow for different structure conventions and representations. If answers are too numerous, then SciFinder again guides the searcher through analysis tools for substances.
This text is not designed to explain the mechanics of searching or of data processing. These are detailed in the numerous help messages available through the help icons in SciFinder.
Instead this text explains what ‘goes on behind the scenes’ and gives examples of search strategies. The text starts with an outline of the basic content of the databases and the way SciFinder searches these databases. It explains why certain answers are retrieved and how features of SciFinder may be used to narrow or broaden searches in a predictable way. It explores different options to the solutions of problems, and above all it encourages scientists to be creative and to think carefully about how to approach problems. SciFinder is a research tool and not just a search tool!
The searches in this text were conducted in the early part of 2009, and the SciFinder functions and types of answers obtained were current at that time.
SciFinder: Setting the Scene Databases in SciFinder Explore by Research Topic
Explore by Chemical Substance
Substructure and Similarity Searching
Additional Search and Display Options
Searching for Information on Chemical Reactions
Some SciFinder Resources
CAS Roles in CAPLUS
Some Basic Principles Used by SciFinder in the Interpretation of a Research Topic Query
Registration of Substances Understanding Structure Searches
Original Publication Discussed in Chapter 7, Section 7.1