Professional Documents
Culture Documents
Data Cleaning
Data Selection
Data Transformation
Data Integration
Data Mining
4
Pattern Evaluation
Knowledge presentation.
Text preprocessing
Text transformation
Attribute selection
Pattern discovery
Interpretation or evaluation
There are additional useful goals for NLP; most of them are related
to the specific application for which it is being exploiting. The aim of NLP
systems is to describe the precise meaning and intension of the user query that
is given in the regular language of the user. Moreover, the contents of the
documents that are being investigated will be represented at all their levels of
meaning so that a true match between need and reply can be found, in spite of
how they are represented in their surface form.
If any application that make use of text is candidate for NLP. The
following are some of the list of common Applications in NLP. It includes, it
tends to have direct real-world applications, while, and more commonly serve
as sub-tasks that are used to aid in solving larger tasks.
NLP tools are used to enhance the quality of retrieval process. ADS
uses the NLP tools and follows the same procedure as if in Informational
Retrieval systems. The most important criteria in ADS is the extracting the
sentences which can be performed by NLP tools such as co-reference
resolution, discourse analysis or named entity recognition.
The most common NLP tools used in ADS are light stemmers, root
stemmers, standard stopwords, domain specific stopwords, parser, Parts-Of-
Speech(POS) Tagger, Word segmentation, Sentence breaking. NLP tools can
integrate with the other model to provide the efficient tools for the ADS,
information retrieval and information extraction. The tools are often used to
perform redundancy elimination, to find relationships and similarities
9
are query biased, they do not offer an overall sense of the document content,
and therefore, are not suitable for substance summary. A query-relevant
summary is prejudiced in favor of a specific question or subject. The query-
based summarization furnishes the summaries intimately linked to the query.
A query-focused summary offers the data that is most pertinent to the
specified queries. When compared to generic summarization, which must
comprise the core data vital to the original documents, the vital aim of query-
focused Document summarization is to generate from the documents a
summary that is capable of answering the requirement for data given in the
subject or explaining the subject.
S Re t S Re l
Re call
S Re t (4.17)
Where, Sret and Srel are the number of retrieved and relevant sentences
respectively.
S Re t S Re l
Pr ecision (4.18)
S Re l
2 Re call Pr ecision
F measure
Re call Pr ecision
INPUT DOCUMENTS
PREPROCESSING
SUMMARIZATION PROCESS
SUMMARY
22
Preprocessing:
Summarization Process
wide scope of the input document set and (ii) diversity, avails priority to
incorporate the dissimilar sentence from the input document set.
Chapter Four focus the deep learning algorithm integrated with fuzzy for
multi-document text summarization
Chapter Five discusses the deep learning algorithm integrated with fuzzy and
genetic particle swam optimization for multi-document text summarization
Chapter Six concludes the finding of this research work and provides future
direction of research.
1.11 SUMMARY