Inverted file

Fig. 4.5 Inverted file of fragments contains identification numbers of all reference structures having the same fragment in the same record. Decomposition of the query structure into fragments and scanning the inverted file yields a short file of possible candidates.

Such tables are very convenient for peak-by-peak search if the inverted files containing ID numbers of reference spectra are at hand. These files must be generated in advance (Fig. 4.7). 

The second problem inherently associated with peak search in the inverted file of peak vs. ID numbers is the tolerance limit within such a retrieval should be carried out. If the intervals in which the peaks are inverted are broad the search will probably yield the correct answer but the list of produced matches will be rather... 

Fig. 4.7 Inverted file for retrieval of infrared spectra generated from pealc tables for fast searching by peak positions (in the tolerance region +10 cm"1).

L17. Lytle, F. E., Computerised searching of inverted files (of infrar-red spectra). Anal. Chem. 42, 355-357 (1970). 

Although both the search terms and the data-base terms can be put into alphabetic order, normally only one is in fact done. This process is called, inversion and the ordered file is known as an inverted file . The other list is then processed serially, from beginning to end, dining the matching process, and is therefore known as a serial file . 

Apart from the use of vocabulary control, there are many systems which aid retrieval by classification. (Indeed, vocabulary control is only one form of classification). It has already been mentioned that the secondary services commonly classify and rearrange the documents they cover. These classifications are frequently computerised and so can be used for computer retrieval. For example, the Universal Decimal Classification 44) is in use as a basis of computerised systems (e. g. 45)j. Much work is also going on in the development of automatic classifications, based on cluster analysis techniques 46,47). In these cases the classification may be applied by assigning the appropriate codes to file items stored in some entirely different order (e. g. chronologically or alphabetically by source journal). Alternatively, the classification scheme may be used as a basis for organising the file, so that all file items falling into a particular class occur together. This is most commonly achieved by the use of the inverted file approach discussed in Section II. 

The inverted file is. then, merely a computer application of traditional indexing philosophy. Perhaps the most important difference between computer searching and manual search of a printed index is that the computer enables coordination to be done much more easily at search time. In a computer search, the required term coordinations can be specified, as a Boolean expression for example, much more precisely than could be achieved by any precoordinated indexing system. Computer systems therefore convey a much greater flexibility. 

Kent, A. K. The UKCIS Inverted File Information Retrieval System, INFIRS. UKCIS Internal Report (1970). 

What these systems have in common is that they attempt to put a large part of the computationally-intensive part of substructure searching into a pre-processing of the database, generating a tree-structured set of inverted files. The search itself can then be relatively rapid. 

Over the past several years work has been underway on the second generation parallel structure search architecture. This architecture is a completely new hardware and software implementation which builds on the already proven architectural principles of the first generation system as well as taking advantage of inverted file searching rather than sequential searching to implement the screen search algorithm. 

CAS has developed a state-of-the-art inverted file search capability in support of text searching and numeric file searching that can readily beapplied to structure screen searching. 

Screen searching is done serially rather than using a more traditional inverted file approach. That is, the entire screen file for a particular file segment is read sequentially from disk into memory. All sets of query screens on the current list of queries to be searched are compared against each file structure screen set. If a query screen set matches a file screen set, then the candidate file structure is passed on to iterative search. 

The search engine architecture will substitute inverted file searching for sequential screen searching. This will permit the structure screen set size to be increased because new screens simply become new terms in the inverted dictionary. It will also permit other numeric or textual information associated with a structure to be added to the inverted screen file and processed as part of the inverted screen searching process. 

Then, the Chemical Compound Index gives a direct file for asking what kinds of molecular properties are discussed for the particular compounds. On the other hand, the Contents gives the inverted file, as via its entries all compounds may be found for which a definte molecular property is listed. 

The new programs make use of inverted file storage on magnetic discs. The initial cost of setting up such files is quite high, but searches can be made quickly and more cheaply. 

In batch processes, these search formulations are keypunched on to cards and read into the computer. An editing process checks the validity of the terms against a dictionary the search process then retrieves a list of relevant references from the inverted file and finally a print program retrieves these references and prints them out in intelligible form. 

P. Willett, Analyt. Chim. Acta, 138, 339 (1982). The Calculation of Inter-Molecular Similarity Coefficients Using an Inverted File Algorithm. 

---