MULTI-DOCUMENT SUMMARIZATION SYSTEM: USING FUZZY LOGIC AND GENETIC ALGORITHM

http://www.iaeme.com/IJARET/index.asp 30 editor@iaeme.com
International Journal of Advanced Research in Engineering and Technology
(IJARET)
Volume 7, Issue 1, Jan-Feb 2016, pp. 30-40, Article ID: IJARET_07_01_004
Available online at
http://www.iaeme.com/IJARET/issues.asp?JType=IJARET&VType=7&IType=1
Journal Impact Factor (2016): 8.8297 (Calculated by GISI) www.jifactor.com
ISSN Print: 0976-6480 and ISSN Online: 0976-6499
© IAEME Publication
___________________________________________________________________________
MULTI-DOCUMENT SUMMARIZATION
SYSTEM: USING FUZZY LOGIC AND
GENETIC ALGORITHM
Ansamma John
Research Scholar, Department of Computer Science
University of Kerala
Kariavattom, Kerala, India
ABSTRACT
In the recent times, the requirement for generation of multi-document
summary has gained a lot of attention among the researchers. Mostly, the text
summarization technique uses the sentence extraction technique where the
salient sentences in the multiple documents are extracted and presented as a
summary. In our proposed system, we have developed a sentence extraction
based automatic multi-document summarization system that employs fuzzy
logic and Genetic Algorithm (GA). At first, the different features are used to
identify the significance of sentences in such a way that, each sentence in the
documents is specified with the feature score. The feature score is then fed to
the fuzzy logic (an AI technique) in which the fuzzy inference engine decides
the importance of the sentences based on the fuzzy rules. The fuzzy rules are
optimized with the help of GA algorithm and the extraction of sentences can
be done based on the fuzzy score of each sentences. A multi document
summary is created from the extracted sentences after removing the redundant
sentences. The experiments have been done using the DUC 2002 dataset and
the summary is evaluated with the measures such as Precision, Recall and F-
measure.
Key words: Multi-document, Summary, Feature, Redundancy, Fuzzy logic,
Fuzzy Rule, Genetic Algorithm (GA), Chromosome, Optimization, DUC 2002
Dataset (Document Understanding Conferences).
Cite this Article: Ansamma John, Multi-Document Summarization System:
Using Fuzzy Logic and Genetic Algorithm. International Journal of Advanced
Research in Engineering and Technology, 7(1), 2016, pp. 30-40.
http://www.iaeme.com/IJARET/issues.asp?JType=IJARET&VType=7&IType=1

Multi-Document Summarization System: Using Fuzzy Logic and Genetic Algorithm
1. INTRODUCTION
The amount of information is getting enlarged day by day, resulting in information
overload. In other words, to utilize the information effectively is a challenging
practical task. An urgent need for text summarization has materialized due to
information overload [1]. Text summarization relates to the process of obtaining a
textual document, obtaining content from it, and providing the necessary content to
the user in a shortened form and in a receptive way to the requirement of user or
application [2]. The technology eases the inconvenience of information overload
because only a concise review has to be considered instead of a complete textual
document [3]. From the early stages of text summarization, its main purpose was to
assist user find information by condensing the vital information from a fundamental
source and providing its shortened form. In this regard, text summarization is
regarded as a mediator between the user and information included in several
documents [4]. However, text summarization is still under research and has so far
dealt with news text. But it is turning out to be a useful tool for information search
and choosing in a diverse media.
Recently, automatic summarization has turned out to be a prominent application.
This is because of the large quantity of information on the Web [8]. Automatic Text
Summarization is a method in which a computer summarizes a text. A text is provided
to the computer and it returns a concise and redundant-less extract of the original text.
Summaries originate from two categories of text sources, a single document or a
document sets [9]. Single document summarization can be defined as the process of
creating a summary from a single text document. Multi-document summarization is
the method of shortening, not just a single document, but a collection of related
documents, into a single summary [10]. Commonly, a precise summary should be
pertinent, short and articulate. In other words, the summary should meet the major
concepts of the original document set, should be redundant-less and ordered [11].
These attributes are the basis of the generation process of the summary. The quality of
summary is sensitive for those attributes relating to how the sentences are scored on
the basis of the employed features. Consequently, the estimation of the efficacy of
each attribute could result the mechanism to distinguish the attributes possessing high
priority and low priority [12].
A multi-document summary possesses some notable merits over a single-
document summary. It offers a domain summary of a topic based on a document set
representing identical information in several documents, distinct information in
separate documents, and association between sections of information across various
documents. It can enable the user to look in for more information on certain facets of
interest, and look into the distinctive single-document summaries [10]. Most of the
similar techniques employed in single-document summarization are also employed in
multi-document summarization. There exist some notable disparities [13]: (1) The
degree of redundancy contained in a group of topically-related articles is considerably
greater than the redundancy degree within an article, since each article is appropriate
to illustrate the most important point and also the required shared background. So,
anti-redundancy methods play a vital role. (2) The compression ratio (that is the
summary size with regard to the size of the document set) will considerably be lesser
for a vast collection topically related documents than for single document summaries.
When compression demands get intensified, summarization becomes challenging. (3)
The co-reference problem in summarization possesses still bigger challenges for
multi-document than for single-document summarization [14].

Ansamma John
In this paper, we have developed an automatic multi-document summarization
system, which employs fuzzy logic and Genetic algorithm. Here, we have used eight
different features to identify the significance of sentences in such a way that each
sentence in the documents is specified with the feature score. Subsequently, the
feature score is applied to the fuzzy logic, an AI technique, in which the fuzzy
inference engine decides the importance of the sentences based on the fuzzy rules.
The optimized rules generated by the Genetic algorithm are used as fuzzy rules. The
sentences are then extracted based on the fuzzy score of the sentences and the
extracted sentences make up a multi-document summary after removing the redundant
sentences. We have used DUC 2002 dataset to evaluate the summarized results based
on the measures such as Precision, recall and f-measure.
The rest of the paper is organized as follows: The review of related researches is
given in section 2. The proposed automatic summarization system is presented in
section 3. The experimental results and analysis are given in section 4. Finally, the
conclusions are summarized in section 5.
2. REVIEW OF RELATED RESEARCHES
A handful of researches are available in the literature to summarize the multiple
documents. Recently, several researches have been presented a multi-document
summarization system based on Artificial Intelligence techniques. Some of the works
presented in the multi-document summarization are given as follows:
Dragomir R. Radev et al. [15] have presented a multi-document summarizer,
MEAD, which created summaries by employing cluster centroids generated by topic
detection and tracking system. It discussed two techniques, a centroid-based
summarizer, and an evaluation scheme on the grounds of sentence utility and
subsumption. The assessment was subjected to single and also multiple document
summaries. In the end, they elaborated about two user studies that test the models of
multi-document summarization. Marie-Francine Moens et al. [22] have analyzed and
discussed about the technologies for single and multi-document summarization which
can be employed on heterogeneous texts for diverse summarization tasks. They have
attributed the removal of main sentences from the documents, compressing the
sentences to the appropriate content, and identifying redundant content throughout the
sentences.
Fu Lee Wang et al. [16] have presented a multi-document summarization system
to obtain the critical information from terrorism incidents. News articles of a terrorism
happening were arranged into a hierarchical tree structure. Fractal summarization
model was used to produce a summary for all the news stories. Experimental results
proved that the system efficiently extracted the main information for the incident.
Dexi Liu et al.[17] have proposed the multi-document summarizer employing genetic
algorithm-based sentence extraction (SBGA) regards summarization process as an
optimization problem where the optimal summary was selected among a summary
sets created by the conjunction of the original articles sentences. To unravel the NP
hard optimization problem, SBGA employed genetic algorithm, which could select
the optimal summary on global aspect. To enhance the correctness of term frequency,
SBGA used a TFS method, which considered word sense while determining term
frequency. The experiments on DUC04 data proved that their strategy was efficient
and the ROUGE-1 score was only 0.55% lesser than the best one in DUC04.

3. A SYSTEM FOR MULTI-DOCUMENT SUMMARIZATION
BASED ON FUZZY LOGIC AND GENETIC ALGORITHM
Multi-document summarization is an automatic process that aims to extract the
relevant summary from the multiple documents that are written about the same
events. The automated procedures for generation of single document summary have
been introduced in 1950's but still it has been received considerable attention among
the researchers. Since the contents on the web are growing very rapidly, there is a
strong need to summarize a large set of documents in a short period of time. So,
several researchers [22-27] have been successfully made use of automated procedure
for generating a relevant, concise and fluent summary from the multiple documents.
In this research, we have developed an automated multi-document summarization
system, which utilizes the fuzzy logic based summarization where the rules are
optimized by the Genetic algorithm.
The steps involved in the proposed system for producing a multi-document
summary from the multiple documents are,
1. Preprocessing
2. Computation of feature score using different features
3. Fuzzy modeling and generation of fuzzy rules using GA
4. Removal of Redundant Sentences
3.1. Preprocessing
The multiple documents are preprocessed using the techniques namely, Sentence
segmentation, removing stop words and stemming. Thus, each sentence with their
corresponding ID and the words containing in each sentence are extracted.
 Sentence segmentation: Separation of each sentences using the delimiter (“.” full
stop).
 Stopword Removal: Removes of stop (linking) words like “have”, “been”, "it", "can”,
“may", "and", "by", "from", "of", "the", "to", "with" and the like from the document
[19].
 Stemming algorithm: Removes the prefixes and suffixes of each word [18].
3.2. Computation of Feature Score Using Different Features
The preprocessed documents are then utilized to compute the feature score for every
sentence in accordance with the eight different features. The different features taken
for our proposed system are as follows:
1. Word similarity among documents: A sentence is assigned by a high score based
on the similar terms (words) among all the documents and the high frequency
count. Here, we take the top n-frequent words from every documents and
identify the similar keywords in the frequent word list. For every sentence, we
count the number of occurrences of these similar identified keywords. Feature
score, word similarity among document is calculated by the ratio of the similar
keywords count of the given sentences to the number of frequent words ( n ) is
taken to find the similarity.
2. Centroid value: Centroid [15] is a feature value which is used to identify the
salient sentences for summarizing the multiple documents. The centroid value for
each sentence ( sC ) is the sum of the individual centroid of the words ( wC )
containing in the sentence. The centroid of each word or term is the product
value of term frequency (TF ) and the inverse document frequency ( IDF ). The

Ansamma John
term frequency (TF ) is the number of occurrences of a given term appeared in
the document. The inverse document frequency ( IDF ) is obtained by the
division of the number of documents in the document set and the number of
documents containing the given word, and then find the logarithm of that
quotient.


m
i
ws i
CC
1
IDFTFCw 







wordgiventhecontaingdocumetsofNumber
documentsofNumber
IDF log
Where, sC  Centroid value of the sentence
wC  Centroid value of the word
TF  Term Frequency
IDF  Inverse Document Frequency
3. Paragraph Frequency: The words are extracted from individual paragraphs and
also identify the number of occurrences of the extracted words among the
paragraphs. The number of occurrences of keywords is summed up to get the
paragraph frequency of the given paragraph. The feature score is computed by the
ratio of the paragraph frequency of the given paragraph to the maximum
paragraph frequency in the document. The feature score for each sentence in the
given paragraph is same as the feature score of the given paragraph.
4. Positional based score: The maximized centroid value is given as a score value
for the first sentence in the given document and the remaining sentences get the
score value based on their corresponding position in the document. The positional
score ksP for all sentences within a document calculated is given in [15]:
max
1
C
n
kn
Pks 


Where, maxC  Maximized centroid value
n  Number of sentences in a document
5. Format based score: In general, the important words of the sentences are
represented with the specific formats like Italics, Bold, underlined and different
font sizes. The words represented with the above specific formats are likely
important for the summarized result. The feature score is the ratio of the number
of words in the sentence with special format to the total number of words in the
sentence.
6. Numerical data: The numerical data presented in the document have some
significant information and it would more likely include in the summary. The
sentences contained numerical data are the most preferable sentences. The feature
score is calculated by the ratio of the number of numerical data occurred in the
sentence and the length of the sentence.

7. Title features: The title words are probably an important feature when
summarizing the document. The feature score for this feature is computed as the
ratio of the number of similar title words in the sentence to the total number of
words in the title.
8. Sentence length: This feature is useful for removing the short sentences.
Generally, the short sentences are the news article contains datelines and author
name. But, it is not necessary for the summary. So, the sentence length feature
has given more priority for the long length sentences rather than the short one. It
is defined as the ratio of number of words occurring in the sentences and the
number of words in the longest sentence in the document.
3.3. Fuzzy Modeling and Generation of Fuzzy Rules Using GA
The term "fuzzy logic" resulted in the development of the theory of fuzzy sets by
Zadeh [20]. The fuzzy logic is extension of the classical logic in form of
generalization of the classical logic inference rules which has capability to deal with
approximate reasoning. The fuzzy set is an expansion for the traditional set “crisp set”
in which each member has a degree of membership to that set determined by
membership function. The membership function is a function that gives membership
degree to each member in the target set, the range of membership degree between
zero and one. The fuzzy logic has benefit in terms of simplicity of development and
modification because the rules are well understandable and simple to modify, add
novel rules or remove existing rules [21].
Here, we make use of the fuzzy logic system in the proposed multi-document
summarization system. The sentences presented in the documents are assigned by a
feature score using the aforementioned eight features. The feature score of each
sentence are given as an input to the fuzzy logic system. The advantages of using
fuzzy logic system are (1) Allows imprecise/contradictory inputs, (2) Permits fuzzy
thresholds, (3) Reconciles conflicting objectives (4) Rule base or fuzzy sets easily
modified. The fuzzy logic system consists of four components: (a) Fuzzifier (b) Rule
base (c) inference engine (c) Defuzzifier.
3.3.1. Fuzzifier
The obtained feature score of every sentence is fed to the fuzzifier, which converts the
numerical data into the linguistic values (High, medium, low). The linguistic values of
each feature score is obtained using the membership function, which is a curve that
defines how each point in the input feature score is converted into a membership
value (or degree of membership) between 0 and 1. Here, we use the triangle
membership function which is defined as follows,
















cxbif
ab
xa
bxaif
ab
ax
axif
cbaxf
,
,
cx,,0
),,:(
Where a, b and c are characteristic parameters of a fuzzy set.
3.3.2 Fuzzy Rule Base
Once the inputs have been fuzzified, we define the fuzzy rules which are important
for any fuzzy system. In general, fuzzy logic based summarization system has used

Ansamma John
the manually defined fuzzy rules. This approach is not an effective one when the
number features and linguistic variable is large. In order to obtain the effective rules,
we have used Genetic algorithm to provide optimized rules for the fuzzy system. The
obtained optimized rules are then stored in the fuzzy rule base.
3.3.3 Generation of Fuzzy Rules Using Genetic Algorithm
Genetic algorithm is used in various applications to find the optimal solution. Genetic
algorithm (GA) is an evolutionary algorithm that evolves computer programs and
predicts mathematical models from experimental data. GA starts with a random
population of candidate solutions in the form of chromosomes. The chromosomes are
then evaluated based on a fitness value and chosen by fitness to reproduce with
modification via genetic operations such as crossover and mutation. The new
generation of solutions goes through the same process until the termination criteria is
satisfied. The fittest individual serves as the final solution. We have utilized the GA
algorithm to find the optimal rules for fuzzy system to summarize the multiple
documents.
 Chromosome initialization: Initially, the set of chromosomes are generated randomly
and each chromosome consists of eight genes that represent the linguistic variable of
each features.
 Fitness function: Fitness function is used to evaluate the survival of the chromosomes
and the fitness function is computed as follows,





MLHJ
MLHJ
m
T
JG
JG
F
,,
,,
)(
)(
; J= {High, Medium, Low}
Where, TF  Fitness function
)(JGm  Number of genes co-occurring in chosen chromosome and the reference
chromosome
)(JG  Number of genes presented in the reference chromosome.
 Crossover and Mutation: In the first level of iteration, two chromosomes are
generated randomly and the fitness value is calculated for the randomly generated
two chromosomes. The crossover and mutation operation is applied over the
randomly generated chromosomes. Here, we used the single point cross over and
mutation operation. Then, the fitness value is computed for the newly generated
chromosomes and the better chromosome is selected from the first iteration. The
better chromosome from the first iteration is to be given to the next level of iteration
with one randomly generated chromosome. Again, crossover and mutation is done
with the above two chromosomes and the better one is selected based on the fitness
value. This process is repeated for ‘n’ number of iterations.
 (d) Chromosome selection: We obtained the better set of chromosomes once the ‘ n ’
number of iteration gets terminated and they are sorted based on the fitness value.
The optimized chromosomes are taken from the sorted list of chromosomes. Each
optimized chromosome represents the fuzzy rule, which is used for finding the
importance of sentences.

3.3.4. Inference Engine
The inference engine is used to provide the score for every sentence based on the
fuzzy rules enclosed in the fuzzy rule base. It provides the fuzzy set {important,
unimportant}, which means whether the sentences are important or unimportant. The
fuzzy rule enclosed in the rule base consists of two parts antecedent and consequent.
Antecedent is defined by the multivariate membership functions and consequent is the
inference of the rule which determines whether the sentence is important or
unimportant based on the input. The fuzzy rule is expressed as: IF (Word similarity
among documents is H) and (Centroid value is H) and (Positional based score is H)
and (Paragraph Frequency is M) and (Format based score is L) and (Numerical data is
H) and (Title features is M) and (Sentence length is L) THEN (Sentence is important).
3.3.5. Defuzzifier
The input for the defuzzifier is a fuzzy set {important, unimportant} and the output is
a crisp value. The resultant crisp value is the final score of each sentence.
3.4 REMOVAL OF REDUNDANT SENTENCES
The sentences that are extracted from the multiple documents based on the final score
generated by defuzzifier. The extracted sentences constitute the summary which may
consist of redundant sentences, due to the fact that the multiple documents are written
about the same events. So it is necessary that the redundant sentences must be
removed from the summary. The redundant sentences are identified by the similar
words occurring among the sentences. The formula used to perform the redundant
removal process in between two sentences is given in [15],
21
2
NN
N
R s



Where, R  Redundancy of two sentences
sN  Number of similar words in the first and second sentences
1N  Number of words in the first sentence
2N  Number of words in the second sentence
4. EXPERIMENTAL RESULTS AND ANALYSIS
This section describes the experimental results of the proposed multi-document
summarization system. The proposed system is implemented in MATLAB
(Matlab7.8). We have used the DUC 2002 dataset in our experiments to generate the
multi-document summary.
4.1 EVALUATION MEASURE
The proposed multi-document summarization system has utilized the evaluation
measures such as precision, recall and F-measure for validating our proposed
approach. For evaluation, we have used the summary presented in the dataset. (1)
Precision: It is the ratio of the number of similar sentences in both the summaries and
the number of sentences in the summary generated by the proposed system. (2)
Recall: It is the ratio of the number of similar sentences in both the summaries and the
number of sentences in the summary presented in the dataset. (3) F-measure: F-

Ansamma John
measure is a measure that combines Precision and Recall. It is defined by the
following equation.
RP
RP
Fm


 2
Where, mF  F-measure
P  Precision
R  Recall
4.2 Experimental Results
The experimental results of the proposed multi-document summarization system are
presented in this section. The multi-documents (written about the same topic)
available in the DUC 2002 dataset is used in our proposed multi-document
summarization system that generates the multi-document summary. The generated
summary is evaluated with the summary presented in the dataset using the measures
such as precision, recall and the F-measure. The evaluation measures are computed
for different percentage in summary (with respects to documents size) and the results
are given in table 1. The corresponding graph is shown in figure 2.
Table 1 Evaluation measures for % in summary
% in summary Precision Recall F-measure
40 0.565 0.43 0.488
50 0.517 0.5 0.508
60 0.5 0.533 0.515
Figure 2 Evaluation measures vs. % in summary
5. CONCLUSION
We have developed automatic multi-document summarization system which
incorporates the fuzzy logic and GA algorithm. We have used eight different features
for feature extraction phase. The feature score of the sentences is applied to the fuzzy
logic system in which the fuzzy rules are optimized with the help of Genetic
Algorithm. We have used DUC 2002 dataset to evaluate the summarized results based
0.4
0.45
0.5
0.55
0.6
40 50 60
% in summary
Evaluationmeasure
Precision
Recall
F-measure

on the measures such as Precision, recall and f-measure. The experiment results
showed that the incorporation of fuzzy logic with GA effectively summarize the
multi-documents.
REFERENCES
[1] Barzilay. R., Elhadad, N and McKeown. K. R., “Inferring strategies for
sentence ordering in multidocument news summarization”, Journal of
Artificial Intelligence Research, vol.17, pp. 35–55, 2002.
[2] Inderjeet Mani, “Automatic summarization”, Kluwer Academic Publishers,
2004.
[3] Jen-Yuan Yeh , Hao-Ren K and Wei-Pang Yang , "iSpreadRank: Ranking
sentences for extraction-based summarization using feature weight
propagation in the sentence similarity network", Expert Systems with
Applications, vol. 35, no. 3, pp. 1451-1462, October 2008.
[4] Yatsko V., Shilov S. and Vishniakov T., “A Semi-automatic Text
Summarization System”, In proceedings of the 10th International Сonference
on Speech and Computer, Patras, pp. 283-288, 2005.
[5] Kaustubh Patil and Pavel Brazdil, "SUMGRAPH: Text Summarization Using
Centrality In The Pathfinder Network", International Journal on Computer
Science and Information Systems, vol.2, no.1, pp. 18-32, 2007.
[6] Ladda Suanmali, Naomie Salim and Mohammed Salem Binwahlan, "Fuzzy
Logic Based Method for Improving Text Summarization", International
Journal of Computer Science and Information Security (IJCSIS), vol. 2, no.
1, 2009.
[7] Ladda Suanmali, Naomie Salim and Mohammed Salem Binwahlan,
"Automatic Text Summarization Using Feature Based Fuzzy Extraction", vol.
20, no. 2, pp. 105-115, November 2009.
[8] F. Canan Pembe and Tunga Gungor, “Automated Query-biased and
Structure-preserving Text Summarization on Web Documents”, in
proceedings of the International Symposium on Innovations in Intelligent
Systems and Applications, Istanbul, June 2007.
[9] Liang Zhou, Miruna Ticrea and Eduard Hovy, "Multi-document Biography
Summarization", in Proceedings of Empirical Methods in Natural Language
Processing, 2004.
[10] Shiyan Ou, Christopher S.G. Khoo and Dion H. Goh, "Design and
development of a concept-based multidocument summarization system for
research abstracts", Journal of Information Science, vol. 34 , no. 3, pp. 308-
326 , June 2008.
[11] You Ouyang, Wenji Li and Qin Lu, "An Integrated Multi-document
Summarization Approach based on Word Hierarchical Representation", in
proceedings of the ACL-IJCNLP, singapore, pp. 109–112, 2009.
[12] Mohammed Salem Binwahlan, Naomie Salim and Ladda Suanmali, "Swarm
Based Features Selection for Text Summarization", IJCSNS International
Journal of Computer Science and Network Security, vol. 9, no.1, January
2009.
[13] Jade Goldstein, Vibhu Mittal, Jaime Carbonell and Mark Kantrowitzt,
"Multi-Document Summarization by Sentence Extraction", NAACL-ANLP
2000 Workshop on Automatic summarization, Seattle, Washington, vol. 4,
pp. 40 - 48, 2000.

Ansamma John
[14] Breck Baldwin and Thomas S. Morton, “Dynamic coreference-based
summarization”, in Proceedings of the Third Conference on Empirical
Methods in Natural Language Processing (EMNLP-3), Granada, Spain, June
1998.
[15] Dragomir R. Radev a, Hongyan Jing, Magorzata Sty and Daniel Tam,
"Centroid-based summarization of multiple documents, Information
Processing and Management, vol. 40, no.6, pp. 919–938, 2004.
[16] Fu Lee Wang, Christopher C. Yang and Xiaodong Shi," Multi-document
Summarization for Terrorism Information Extraction”, Lecture Notes in
Computer Science", Springer, vol. 3975, May 2006.
[17] Dexi Liu, Yanxiang He, Donghong Jiand Hua Yang, "Genetic Algorithm
Based Multi-document Summarization”, Lecture Notes in Computer
Science", Springer, vol. 4099, pp. 1140-1144, July 2006.
[18] Lovins, J.B. 1968: "Development of a stemming algorithm". Mechanical
Translation and Computational Linguistics, vol. 11, pp. 22-31, 1968.
[19] Pant, G., Srinivasan, P., Menczer, F., "Crawling the Web", Web Dynamics:
Adapting to Change in Content, Size, Topology and Use, edited by M.
Levene and A. Poulovassilis, Springer- verilog, pp. 153-178, November
2004.
[20] Zadeh, L.A., “Fuzzy sets”, Information and control, vol.8, pp. 338-353, 1965.
[21] Klir, G.J. and B. Yuan, “Fuzzy Sets and Fuzzy Logic: Theory and
Applications”, 1st Edn., Prentice-Hall, New York, pp: 592, 1995.
[22] Marie-Francine Moens,Roxana Angheluta and Jos Dumortier, "Generic
technologies for single- and multi-document summarization", Information
Processing and Management, vol. 41, no.3, pp.569-586, May 2005.
[23] Jiaming Zhanb, Han Tong Lohb and Ying Liua, "On macro- and micro-level
information in multiple documents and its influence on summarization",
International Journal of Information Management, Vol. 29, No. 1, pp. 57-66,
February 2009.
[24] Sugandhi Midha, Comparative Study of Remote Sensing Data Based on
Genetic Algorithm. International Journal of Computer Engineering and
Technology, 5(1), 2014, pp. 141-152.
[25] Palak Sharma, Shelza, Gurjit Singh Randhawa and Rajinder Singh Virk, Cost
Optimization of Pipeline Systems Using Genetic Algorithm, International
Journal of Computer Engineering and Technology, 4(4), 2013, pp. 485-494.
[26] David M. Zajica, Bonnie J. Dorra and Jimmy Linb, "Single-document and
multi-document summarization techniques for email threads using sentence
compression", Information Processing and Management, Vol. 44, No. 4, pp.
1600-1610, July 2008.
[27] Lin Zhao, Lide Wua, and Xuanjing Huang, "Using query expansion in graph-
based approach for query-focused multi-document summarization",
Information Processing and Management, Vol. 45, No. 1, pp. 35-41, January
2009.
[28] Yohei Seki, Koji Eguchi and Noriko Kando, " Multi-Document Viewpoint
Summarization Focused on Facts, Opinion and Knowledge”, Computing
Attitude and Affect in Text: Theory and Applications, Springer, Vol. 20,
pp.317-336, January 2006.
[29] Florian Boudin and Juan Manuel Torres Moreno, " NEO-CORTEX: A
Performant User-Oriented Multi-Document Summarization System”, Lecture
Notes in Computer Science", Springer, Vol 4394, pp.551-562, May 2007.

MULTI-DOCUMENT SUMMARIZATION SYSTEM: USING FUZZY LOGIC AND GENETIC ALGORITHM

More Related Content

What's hot (18)

Similar to MULTI-DOCUMENT SUMMARIZATION SYSTEM: USING FUZZY LOGIC AND GENETIC ALGORITHM (20)

More from IAEME Publication (20)

Recently uploaded (20)

MULTI-DOCUMENT SUMMARIZATION SYSTEM: USING FUZZY LOGIC AND GENETIC ALGORITHM