Você está na página 1de 8

Annals of Library and Information Studies

PANIGRAHI
Vol. 54, March 2007, pp. 37-43 & PRASAD: FACET SEQUENCE IN ANALYTICO SYNTHETIC SCHEME 37

Facet sequence in analytico synthetic scheme: A study for developing


an AI based automatic classification system

Pijushkanti Panigrahia and A R D Prasadb


a
Reader, Department of Library and Information Science, University of Calcutta, 87/1 College Street,
Kolkata, PIN 700 073, West Bengal, E-mail: panigrahipk@yahoo.com

b
Associate Professor, Documentation Research and Training Centre, Indian Statistical Institute, R V College - P O,
Bangalore, PIN - 560 059, E-mail: ard@drtc.isibang.ac.in or ardprasad@hotmail.com

Aims to demonstrate the techniques of fixing the facet sequence in developing an automatic classification system to
construct classification numbers for document titles, which appear in natural language. Describes the procedure of knowledge
representation for basic subjects and their facets with regard to five fundamental categories, their respective notations, symbols
etc. States the techniques of selecting their respective category name to follow the facet formula in general. Discusses the
method of identification of facets from the title of the document. Also includes the scope of defining the specific facet formula
for any specific subject. Facet formula for ACI of Colon Classification is also included to handle it successfully following
postulates and principles.

Introduction
of document titles. Canons, principles, postulates and
Prof S R Ranganathan bestowed an elaborate scientific
different instructions mentioned within schedules are
theory of library classification after a careful study of
implemented to a large extent through an inference
the various ways of formation of new subjects in the
engine developed in Prolog language4. The system is
‘Universe of Subjects’ and introduced the concept of
named as Viswamitra 5.
Analytico Synthetic Classification system devising
Colon Classification (CC) in 19331. These studies have
As the time isolates can appear with any one of the Main
obviously to do a lot with basic ideas and their relations.
Classes enlisted in the schedule of basic subjects, the
Typically ideas have a kind of CoSSCo relationship
schedule for time isolates is mentioned as a common
among themselves. The parallels can be easily seen
schedule in CC7. From the very beginning CC has
between the study of Universe of Knowledge and that
separately specified isolate numbers for time, which is
of Knowledge Based Computer Systems (KBCS)
one of the five fundamental categories. An elaborate
otherwise known as expert system in the field of
study is made to investigate the importance of time (i.e.
Artificial Intelligence (AI). Although both fields have
Meteorological time, Civil time etc) in classifying the
grown independently, they have a lot to offer to each
thought content of any document in the automatic
other. The application of Natural Language Processing
environment with proper implementation of the structure
(NLP), another important area of AI could solve the
of the schedule of the Common Time Isolates including
problem of designing a powerful automatic classification
its all divisions and subdivisions6. The expert system is
system 2.
developed to handle the procedure of picking up numbers
from other schedules and forming the special
An AI based automatic classification system
components to build compound time isolates 7.
An AI based automatic classification system is developed
for the CC7. This system consists of a large number of From the very beginning CC has separately specified
knowledge bases to represent the schedules of common isolate numbers for Space or Geographical areas, which
isolates and special isolates following the frame based is one of the five fundamental categories. As the space
knowledge representation model3. A huge lexicon is isolates can occur in any one of the Main Classes enlisted
developed to support the natural language based analysis in the schedule of basic subjects, the schedule for space
38 ANN LIB INF STU, MARCH 2007

isolates is mentioned as a common schedule. A study is important and must be helpful. In the automated
made to explore the importance of space / geographic environment, facets, their categories and corresponding
area in classifying the thought content of any document notations are stored in a b-tree which follows
in the AI based automated environment with proper alphabetical arrangements as storage technique and this
implementation of the structure of the Common Space is used for future retrieval. A planning for achieving this
Isolates including the political divisions and subdivisions goal is taken by adopting two ways – proper
of continents, subcontinents and the administrative nomenclature of the categories and writing rules for facet
divisions of the country. Frame based knowledge formula. Thus the nomenclatures of the categories are
representation technique is used to develop knowledge to be changed carefully so that their positions are
base for different groups of isolate ideas as derived in properly sequenced in the b-tree according to canons,
CC7 using characteristics like ‘by geographical feature’, postulates and principles. The following subsections are
‘by population cluster’, ‘by zone’, ‘by area surrounded dealt with this direction. Frame based knowledge
by ocean’, ‘by orientation’, ‘by empire’ etc. representation technique is used to develop a Knowledge
Geographical areas, their divisions and subdivisions with base to represent the basic subjects, special isolates,
imposed rules are mentioned at the beginning of the categories of isolates, the isolate numbers, etc.
knowledge base for space isolates; and the knowledge
base for special components to form the compound Basic Subject (B S)
isolates if necessary, are mentioned thereafter8.
An AI system must be domain specific. ‘Medicine’ is
Various devices for Colon Classification are used to considered as the domain for developing an automatic
specify any isolate, especially when a new subject enters classification system. Hence a frame based knowledge
into the universe of subjects. The procedure of building representation model is developed to represent the
numbers for compound isolates using Superimposition semantics of isolate ideas belonging to the basic subject
device, Alphabetical devices, Geographic devices etc. ‘Medicine’. For example, the following Prolog fact may
is tested in the automatic classification system using AI be added to the knowledge base.
techniques.
cc7 ( medicine, number, ’L’ ).
Similarly many other related matters (e.g. System,
Environment, Special constituents, ACI etc) are well
The fact ‘Basic subject medicine has the notation L’ is
tested in the AI based automated environment. But an
true for all special isolates. But this fact is used only
obvious barrier was found while synthesizing facets in
once, and inheritance mechanism can be invoked to find
their helpful sequence. This specific study is performed
out the basic subject of individual isolate idea and its
to overcome this problem.
subsequent as mentioned earlier. Similarly, for any
topmost isolate, the basic subject is to be mentioned only
Objective of the study once. e.g. cc7(digestive^system, basic_subject,
Due to its unique analytico synthetic nature, Colon medicine).
Classification is considered for performing study of
developing an automatic classification system. Some basic subjects are formed either as systems,
Synthesising facets in this regard is very important. Thus environment or special basic subjects, which are
this paper aims to demonstrate the methods of identifying accompanied and/or denoted by systems constituents,
facets and synthesizing them in a helpful sequence environment constituents or special constituents
according to canons, postulates and principles. More respectively. For examples, Child medicine L 9C, Female
specifically this paper tries to find the position of medicine L 9H etc belong to specials category; whereas
different facets in the facet formula in the automated Ayurvedic medicine L B, Homeopathic medicine L L etc
environment. belong to systems category of medicine. CC7 provides
the sequence of these constituents as systems,
Knowledge representation techniques for facets environment and/or special9. To fix their sequential
In CC, facets belong to five fundamental categories position in the facet formula these names have been
mainly i.e. Personality, Matter, Energy, Space and Time. adopted as ‘a_system’ for system constituents,
While building class numbers, their sequence is ‘b_environment’ for environment constituents and
PANIGRAHI & PRASAD: FACET SEQUENCE IN ANALYTICO SYNTHETIC SCHEME 39

‘c_specials’ for special constituents. Again they are concepts by the following Prolog rules developed for
connected using the indicator digit ‘-’(hyphen). They this purpose 11, 12.
are represented in the Prolog as
find_category (Term, Basic_no, Category, Symbol, Number):-
cc7(a_system, symbol, ’†’). has_value(Term, basic_subject, Basic),
cc7(b_environment, symbol, ’†’). has_value(Basic, number, Basic_no),
cc7(c_specials, symbol,’†’). has_value(Term, category, Category),
cc7(Category, symbol, Symbol),
has_value(Term, number, Number).
Personality (P)
has_value (Term, Slot, Value):-
cc7(Term, Slot, Value), !.
The category personality is represented by ‘e_1p1’ (i.e.
has_value (Term, Slot, Value):-
Personality of first round first level) in knowledge base. cc7(Term, part_of, Superconcept),
Actually it is ‘1p1’ as per Colon Classification 7th has_value(Superconcept, Slot, Value).
edition10, but to identify its position easily in the facet
sequence ‘e’ is added. This is to force the isolates of Where, Term is a variable to represent any isolate.
personality to appear before matter-property, which are (Arguments started with uppercase letter are variable
represented by ‘mp’. In the b-tree where the isolates and started with lowercase letter are constant.) The
along with their categories are stored, the personality has_value predicate is to perform inheritance
isolates appear before matter-property isolates as they mechanism, which again in turn helps to find the
are represented by e_1p1 and mp respectively (i.e. in category, symbol and basic number of any isolate (i.e.
the alphabetical order). Similarly, personality of the first Term) for find_category predicate.
round and second level can be represented as ‘e_1p2’.
However, this addition of a character does not in any Matter Property (M)
way affect the system, but on the other hand helps the
program to determine the facet sequence. This technique The matter property of medicine includes broadly the
is adopted only to set the ordinal value of the facets for following isolates
facet formula. The knowledge base contains following 2 Morphology
facts 3 Physiology
4 Disease
cc7 ( e_1p1, symbol, ’,’ ). 5 Personal hygiene etc.
cc7 ( circulatory^system, category, e_1p1 ).
cc7 ( circulatory^system, basic_subject, medicine ). Of theses, the mostly used matter property isolate is
cc7 ( circulatory^system, number, ’3’ ). ‘Disease’ and its sub divisions. Indicator digit for matter
is ’;’ (semicolon)13. This category is represented in the
Here, circulatory^system belongs to the basic subject knowledge base as ‘mp’, so that the isolates of this
‘medicine’, the symbol for Medicine is ‘L’. The category appear after the personality isolates
circulatory system belongs to the category of e_1p1 (represented by e_1p1) 14. This also helps in placing the
(personality), whose indicator digit is ’,’ and the isolate matter-property isolates before energy isolates (i.e.
number is ’3’. ‘r_energy’).
In case of heart, it is a part of circulatory system. This
information gives the impression that all attributes (like Following is the representation in the knowledge base.
basic_subject, category, symbol ) of circulatory system cc7 ( mp, symbol, ’;’ ).
is applicable to ‘heart’. These attributes can be
accumulated by part_of inheritance method. Again ‘disease’ belongs to the basic subject ‘medicine’,
and to the category ‘mp’ and has isolate-number ‘4’. If
cc7 ( heart, part_of, circulatory^system ). disease is an object, then following are the facts in the
cc7 ( heart, number, ’32’). knowledge base
Hence, there is no need to mention the individual facts cc7 (disease, basic_subject, medicine).
that indicate the basic subject, category etc., of the cc7 (disease, category, mp).
isolate t:‘heart’. Those will inherit to subordinate cc7 (disease, number, ’4’).
40 ANN LIB INF STU, MARCH 2007

document in another paper entitled ‘Inference Engine


The indicator digit for mp is ’;’. The part_of presentation for Space Isolates of Colon Classification for AI Based
is used for successive isolates. Automated Classification System’. It also discusses the
structure of the schedule of the Common Space Isolates
Energy (E) including the political divisions and subdivisions of
continents and subcontinents and the administrative
Generally energy manifests as action, reaction or divisions of the country. The procedure of developing
interaction of some kind or the other. Diagnosis, the knowledge base applyng frame based knowledge
preventive measures or steps, treatment, surgery etc., can representation technique for divisions and subdivisions
be considered as action on disease. So they belong to the of geographical areas, for different groups of isolate
energy category. It is mentioned as ‘r_energy’. According ideas as derived in CC7 using characteristics like ‘by
to Ranganathan, the energy (E) is supposed to appear in geographical feature’, ‘by population cluster’, ‘by
between Matter, Property and Space. So, ‘r’ is preceded zone’, ‘by area surrounded by ocean’, ‘by orientation’,
the term energy to fix its position in the facet sequence. ‘by empire’ etc. are mentioned there clearly. Rules of
Thus ‘r_energy’ is used to fix its position before ‘space’. expert systems developed using Prolog to handle the
procedure of picking up numbers from other schedules
The indicator digit for it is ‘:’(colon) and the and forming the special components to build compound
corresponding fact is represented as space isolates are discussed in detail 16.
cc7 (r_energy, symbol, ’:’).
Time (T)
‘Nursing’ is an isolate, which belongs to Energy. So
related Prolog facts are Time is one of the easily identifiable manifestations.
cc7 (nursing, basic_subject, medicine). From the very beginning CC has separately specified
cc7 (nursing, category, r_energy). isolate numbers for time, which is one of the five
cc7 (nursing, number, ’1’). fundamental categories. As the time isolates can appear
with any one of the Main Classes enlisted in the schedule
The system gathers all information about nursing using of basic subjects, the schedule for time isolates is
the above slots. mentioned as a common schedule. And since the time
isolate of Colon Classification (CC), 7th edition appears
In CC7, the isolate idea ‘therapeutics’ is used as standard after the space facet, so ‘time’ could be used as category
term instead of ‘treatment’. But query language may name in the automatic environment. The importance of
contain ‘treatment’ instead of ‘therapeutics’. So a time (i.e. Meteorological time, Civil time etc) in
thesaurus entry can be used as Prolog fact classifying the thought content of any document has
cc7 (treatment, use, therapeutics). been described in detailed in another paper entitled “A
Knowledge Representation Model for Time Isolates of
The hierarchical relation of any order can be presented Colon Classification (7 th Edition)” 17 including the
with the help of part of relationship 15. structure of the schedule of the Common Time Isolates
including all divisions and subdivisions. Rules of expert
Space (S) systems developed using Prolog to handle the procedure
This category is very easy to identify from their of picking up numbers from other schedules and forming
manifestations. But procedure of its full implementation the special components to build compound time isolates
following the CC7 schedule is very important. Since the have been incorporated for the automated system. Some
Space facet can occur only in the last round among the special techniques of defining lexical knowledge
facets of compound subject, it is sufficient to represent regarding time isolates are also adopted18.
this by ‘space’. As the space isolates can occur in any
one of the Main Classes enlisted in the schedule of basic Anteriorising Common Isolates (ACI)
subjects, the schedule for space isolates is mentioned as Some documents are considered as ‘Approach
a common schedule. The importance of space / Documents’ as user may approach for any one of those
geographic area in classifying the thought content of any ‘Form of Expositions’ to have a look on them before
PANIGRAHI & PRASAD: FACET SEQUENCE IN ANALYTICO SYNTHETIC SCHEME 41

taking up the study of a normal book or any other cc7(u_aci, symbol, ’”’).
document on any subject. On account of this, Dr S.R.
Ranganathan has called these ‘Form of Expositions’ as Facet Formula
Anteriorising Common Isolate (ACI) and decided that
subjects having an Anteriorising Common Isolate The second method of sequencing facets properly is
should precede the host subject on the shelf. This ACI framing rules for facet formula. There are again two ways
can be attached to a host class number with a connecting of it.
symbol -”- (double inverted comma). Any Host Number
followed by the digit –”– (double inverted comma) will Facet Formula in General
precede – that is, will be anterior to – Host Number 19.
These isolates are used for documents which require to Though as an analytico synthetic scheme CC7 does not
be arranged anterior to the ordinary documents on the have rigidity in facet formula, the general facet sequence
concerned subject. In CC7, bibliography, encyclopedia, which is based on the postulate, may be adopted. In most
concordance, etc., are treated as ACI. On the basis of cases, facet formula may follow the P, M, E, S, T
position of their attachment with the host subject three sequence. The basic subject precedes personality (P),
different ACIs have been recognized. They are (i) Matter property (M), Energy (E) with their respective
applicable before space facet, (ii) Applicable only after rounds and levels followed by Space (S) and Time (T).
Space Facet and (iii) Applicable only after Time Facet The individual categories must be identified with their
as suggested in CC7 20, 21. Each has its own facet formula corresponding indicator digits. The simplest form of a
which may be used for further individualization of these facet formula would look like23
isolates. As far as the knowledge representation model
is concerned, a special technique for presenting (BS) , [P] ; [M] : [E] . [S] ‘ [T]
knowledge is adopted. Thus, the categories of ACI
are identified and named as In Viswamitra, a b tree data structure is used for storing
information/data in the computer memory. Prolog
ACI applicable before space aci1 s_aci compiler does this using ASCII value of characters. To
get the sorting effect, the names of categories have been
ACI applicable only after space aci2 t_aci changed to match the sequence depending on the ASCII
values as discussed in the earlier section.
ACI applicable only after time aci3 u_aci
Hence, constituents to basic subject (e.g. system,
The first character in the category-names has no specific environment and specials), and isolate facets (i.e
connotation, but they are chosen so to get the sorting personality, matter property, energy in different rounds
order only. For example, and levels) and space, time will appear. For fixing their
positions, following names are taken so that they will
s_aci appears before space ( as 95( i.e. ASCII value
of the under score _ ) < 112 ( i.e. ASCII value appear sequentially in the b-tree24.
of p)
t_aci appears after space and before time ‘a_system’ for system constituents
u_aci appears after time
‘b_environment’ for environment constituents
Due to the ASCII values of characters they would appear ‘c_specials’ for special constituents.
in the proper sequence. ‘e_1p1’ is used to represent personality of first round,
first level.
In the knowledge base, the following facts indicate that
they all use the same symbol 22. ‘mp’ is used to represent matter property
‘r_energy’ is used to represent energy.
cc7(s_aci, symbol, ’”’). ‘space’ is used to represent space.
cc7(t_aci, symbol, ’”’). ‘time’ is used to represent time
42 ANN LIB INF STU, MARCH 2007

Respective indicator digits can be incorporated in the The formula predicate is supported by another predicate
knowledge base through the following facts, isitover, which checks the presence or absence of any
category.
cc7(a_system, symbol, ’ ’). isitover ( C, L ) :
cc7(b_environment, symbol, ’ ’). not(lmember(C, L)), !.
cc7(c_specials, symbol,’ ’).
isitover(C,[[]|_]): !.
cc7(e_1p1, symbol, ’,’ ).
cc7(mp, symbol, ’;’ ). isitover (C, [[C,_] | T)): !.
cc7(r_energy, symbol, ’:’ ). isitover (C, [[Cat, Num] | T]):-
cc7(space, symbol, ’.’ ). recorded(done, Cat+N, _),
cc7(time, symbol, ’ ‘‘’ ). isitover(C,T).

However, the above-mentioned names are used to isitover(C, [[Cat, Num] | T]):
maintain facet sequence while constructing the class recorded (out,Output,_),
number. Names for categories of different rounds and cc7(Cat, symbol, S),
levels are fixed in such a way that their position is write(S), write(Num),
adjusted. write(Output,S),
write(Output,Num),
Facet Formula in Specific record(done, Cat+Num, _),
isitover(C, T).
Viswamitra allows defining specific facet formula for
isitover(_,_).
different subjects. Specific rules can also be
implemented in constructing a class number. For
The first isitover(C, T) predicate is used to check the
example, in case of Medicine, CC7 suggests that
list membership using lmember predicate. If it is found
‘disease’ cannot appear unless an organ of human body
that C (i.e. category) is not a member of L (i.e. List)
appears (i.e. Wall Picture Principle) 25 . Similarly
then the system proceeds further and uses other isitover
‘treatment’ has no meaning, if the concept of ‘disease’
predicate to pick up the category, symbol and its isolate
does not exist. In an expressive title, if no organ is
number as defined in the facet formula27.
mentioned but only disease is mentioned, the system
should pick up the number for ‘body and its organ’, i.e.
Conclusion
1 at least and fill the position for personality (e_1p1).
So is the case with the isolate ‘Treatment’.
Viswamitra is an attempt to build an automatic
classification using the tools and techniques of Artificial
A facet formula is defined for medicine in the knowledge
Intelligence (AI), especially Natural Language
base following the list data structure of Prolog26
Processing (NLP) and Expert System (ES). The system
cc7 ( medicine, formula, [[e_1p1, ’1’],[mp,’4’], comprises of two major components viz, the NLP
[r_energy, ’6’] ] ). component and the knowledgebase modeled on frame
based ES. These two components compliment each other.
Here, the third argument is a Prolog list structure of The purpose of the system is to construct classification
categories and their corresponding isolate numbers, numbers using CC7 schedule, which is represented in
which will represent that category as a minimum the form of knowledge base.
requirement in the classification number. If any of these The system has proven the fact that Ranganathan’s idea
three categories are absent, the corresponding isolate of defining isolates in the line of meccano set is the best
number, which appears next to the category name in the suited and highly computer compatible to develop an
facet formula, is picked up. To perform this job the automated classification system using artificial
inference engine has following rules. intelligence techniques. Viswamitra uses the frame-
based expert system. In a frame-based system,
formula (C+N): description about any isolate idea is represented in the
cc7 ( medicine, formula, L ), !, form of a frame. And it becomes easier to identify the
isitover(C,L).
isolate idea, its number, its category, its basic subjects
PANIGRAHI & PRASAD: FACET SEQUENCE IN ANALYTICO SYNTHETIC SCHEME 43

etc., from this model. The NLP components help in facet 11. Panigrahi P, Design and development of a knowledge
analysis. The system is able to follow the facet formula representation model for analytico synthetic classification. Ph
D Thesis. Vidyasagar University, Medinipur (West Bengal,
being guided by principles of helpful sequence for facets. India). (Unpublished), (1996) p72.
Flexibility in it also can be imposed. Finally synthesing 12. Panigrahi P, Logic programming for automatic classification,
the classification number becomes easier. IASLIC Bulletin, 48 (2) (2003) 83-98.
13. Ranganathan S R, Colon classification. 7th ed (Reprint), (ed
References by M A Gopinath). Bangalore: Sarada Ranganathan Endowment
for Library Science. (1988) p30
1. Gopinath M A, Colon classification, edition 7 : theory and
practice. In Seventeenth DRTC Refresher Seminar, Oct 26-28, 14. Op cit. 3
1988. Documentation Research and Training Centre and Sarada 15. Op. cit. 5
Ranganathan Endowment for Library Science, Bangalore, 16. Op. cit. 8
Documentation Research and Training Centre: Bangalore. 17. Op. cit. 7
1988, p1. 18. Op. cit. 6
2. Panigrahi P, Design and development of a Knowledge 19. Panigrahi P and Prasad A R D, Inference engine for devices of
Representation Model for Analytico Synthetic Classification. colon classification in AI based automated classification system,
Ph D Thesis. Vidyasagar University, Medinipur (West Bengal, Annals of Library and Information Studies, 52 (4) (2005) 124-
India). 1996, (Unpublished) 134.
3. Panigrahi P, An Artificial Intelligence approach towards 20. Ranganathan S R, Colon Classification. 7th ed (Reprint), (ed
automatic classification (Part 1), IASLIC Bulletin, 45(2) (2000) by M A Gopinath). Bangalore: Sarada Ranganathan Endowment
73-81. for Library Science. (1989) p33
4. Panigrahi P, An Artificial Intelligence approach towards 21. Ranganathan S R, Colon Classification. 7th ed (Reprint), (ed
automatic classification (Part 2), IASLIC Bulletin, 45(3) (2000) by M A Gopinath). Bangalore: Sarada Ranganathan Endowment
105-118. for Library Science. (1989) p105
5. Panigrahi P, Prasad A R D and Basu A, NLP based automatic 22. Panigrahi P, Developing an intelligent system for library works
classification system for analytico synthetic scheme, SRELS and services using declarative language: a case study of library
Journal of Information Management, 40 (4) (2003) 289-312. classification, CLIS Observer, 20 (3&4) (2003) 57-71.
6. Prasad A R D and Panigrahi Pijushkanti (1997) An inference 23. Gopinath M A, Colon classification, edition 7 : theory and
engine for time isolates of colon classification schedule. practice. In Seventeenth DRTC Refresher Seminar, Oct 26-28,
International Study Conference on Classification Research, 6th, 1988. Documentation Research and Training Centre and Sarada
London. FID/CR and University College London, June 16- Ranganathan Endowment for Library Science, Bangalore,
18, 1997. London. Documentation Research and Training Centre: Bangalore,
7. Prasad A R D and Panigrahi P, A knowledge representation (1988) .
model for time isolates of colon classification (7th Edition), 24. Panigrahi P, Knowledge Representation Models. Proceedings
SRELS Journal of Information Management, 42 (3) (2005) of the Retraining of University and College Teachers under
221-244. the U G C Sponsored University with potential for Excellence
8. Panigrahi P and Prasad A R D, Inference engine for space Scheme on Knowledge Management Organized by the
isolates of colon classification for AI based automated Department of Library and Information Science, Jadavpur
classification system, SRELS Journal of Information University, Kolkata during January 12-13, 2006
Management, 42 (4) (2005) 383-408. 25. Ranganathan S R, Prolegomena to Library Classification. 3rd
9. Gopinath M A, Colon classification, edition 7 : theory and ed. (Reprinted in 1989). Bangalore: Sarada Ranganathan
practice. In Seventeenth DRTC Refresher Seminar, Oct 26-28, Endowment for Library Science.
1988. Documentation Research and Training Centre and Sarada 26. Clocksin E J and Mellish C S, Programming in Prolog. 3 rd
Ranganathan Endowment for Library Science, Bangalore, rev ext edn,. (Narosa Publishing House, New Delhi), 1993.
Documentation Research and Training Centre: Bangalore.
27. Panigrahi, Pijushkanti (1996), Design and development of a
1988, p14
Knowledge Representation Model for Analytico Synthetic
10. Ranganathan S R, Colon classification edition 7(1971): a Classification. Ph D Thesis. Vidyasagar University, Medinipur
preview, Library Science with a Slant to Documentation, 6 (3) (West Bengal, India). (Unpublished).p231-6
(1969) Paper M
66
44 ANN LIB INF STU, MARCH 2007

Corrigendum

It is regretted that a wrong figure was inserted above the caption, “Figure 4 – Bradford – Zipf
Bibliograph for A J Legget” on page 208 in Annals of Library and Information Studies, 53 (4)
2006. The correct figure is as follows:
FREQUENCY OF NUMBER OF PUBLICATIONS

CUMULATIVE NUMBER OF PUBLICATIONS


FREQUENCY
50
CUMULATIVE
140
45

40 120

35
100
30
80
25

20 60

15 40
10
20
5
0
0
0 5 10 15 20 25 30 35 40 45 50

CHANNEL NUMER

Figure 4 — Bradford – Zipf Bibliograph for A J Leggett

— Editor

Você também pode gostar