Text mining on criminal documents

International Journal of Advances in Electronics and Computer Science, ISSN: 2393-2835 Volume-3, Issue-9, Sep.-2016
Text Mining on Criminal Documents
28
TEXT MINING ON CRIMINAL DOCUMENTS
1
RIYA, 2
NAMITA GANDOTRA
1,2
Shoolini Univerity, Solan, H.P.
E-mail: 1
mrsmalhans@gmail.com, 2
namita.saini13@gmail.com
Abstract— Criminal cases that has similar facets means that case has many faces that may receive very different punishment
in court sentence. Court system faces severe challenges on this problem, while search for similar criminal cases are manually
from historical judgments database. It is very time consuming and it is not possible for a judge to do so for most of the
criminal cases. This paper tries to use text mining technology to provide best solution for the criminal cases.
Index Terms— Criminal Case Judgments, (Referential Prison Term Generation, Judge’s Sentence.
I. INTRODUCTION
The main idea of this paper is used for police
investigation document of criminal case send from
district police office , parse this document into
keyword table, then match this data with the trained
database in the form of key word table format of
court’s judgment. Using Cosine similarity algorithm
of text mining technology to calculate coefficient of
similarity, base on highest coefficient, we will find the
closest judgment of the criminal case.
Text mining is also called as data text mining which
is equal to text analytics which means deriving high
quality information from raw data or raw text. It can
be done by devising of patterns and trends through
pattern learning. Text mining involves process of
structuring the input data along some derived
linguistic features and removal of others un-useful
data or meaning-less data and then that data is
inserted into the database and then deriving data with
the structured data and the evaluation and the
interpretation of the output.
Main Idea of This Research
The Procedure of criminal case handling in court.
Procedure of drug abuse criminal case in court, most
and large, start from receiving a Indictment send by
District prosecutor’s office. In indictment, facet of this
criminal case, articles of law applied, list of
evidences, information of offender as well as victims
are described. Judge base on these materials,
testimony of on the court, make his judgment. When
making judgment, judge need to clarify the
consistence of facet and evidences of criminal case,
previous criminal record and if possible, check the
term of punishment that similar case been sentenced,
to decide which articles of criminal law be applied
and to what should extent should the offender be
punished. Security and safety are high priority
concerned issues of civics in a country, court sentence
of criminal cases, focused by press, bring strongly
impact to citizens. People always request for fare
punishment to those criminal case.
Security and safety always rank as high priority
concerned issues in country. People always have a
request for adequate and fare punishment to those
criminal cases, while we may found most and large,
criminal case that has similar faces may received very
different punishment. This is the serious challenge to
court system. To make fare and balanced judgment,
judges must search for similar criminal cases from
historical judgments database, to find out punishments
that had been made before he makes his decision. In
year 2008 each judge of district courts in Ludhiana
and Chandigarh has 75% (average) per month. Now in
year 2015 each judge of district court in LDH have
65% (average) per month, and in High court CHD
each judge have 60% (average) per month complaint
sounds on overloading is laud and clear. Searching for
similar criminal cases manually from historical
judgments database is very time consuming and is it
not possible for a judge to do in most criminal cases
handling except some critical cases. This paper will
tries to use text mining approach to solve problems
faced by judges and people should get more aware
about the rules and punishments so that they can safe
their children from many things.
Text Mining parse unstructured document into
meaningful elements and used to execute further work
as data mining technology do. In this paper we try to
use cosine similarity approach of text mining to find
similar judgments for judge to make fare and balanced
judgment. And govt. also should get aware so that
govt. can make strict rules so that people should fear
from doing anything wrong.
Text mining approach may also be used to solve
another tough issue that judges are suffered. The
quantity of criminal cases in year 2015 in Ludhiana is
up-to 405,357, while 60% of criminal cases are crimes
of drug abuse, public danger, larceny and fraud, these
types of criminal cases are relative simple that cases
of killing, corruption etc., but judges still spend much
of the costly time to handle these cases.
If we classify judgments of drug abuse criminal cases
into several clusters base on the similar facet of
criminal case and assign a judgment template for each
cluster base when we building up training database for
cosine similarity comparison. Then we can use cosine
similarity comparison approach to find the nearest
judgments for a target indictment that send from

29
district prosecutor’s office. We can then find the best
fit judgment template base on which cluster it
belongs, by the way we can digest import stuff from
indictment (feature extraction) and paste it to the
template, a draft of court judgment may generated
automatically. Judge only need to make necessary
modification instead of writing a full judgment
manually.
Constrained by time and resource, this paper will
focus on the first issue, other topic is govt. should get
more alert and make strict and more strict rules so that
nobody can think of doing any wrong work that must
be punishable.
II. OBJECTIVE OF THE STUDY
Based on all the above discussion some of the
objective of my study is as following:
 Collection of data regarding criminal cases.
 Making text files from doc files.
 With the help of database making a decision
supportive.
Criminal case that has similar facets may receive very
different punishment in court sentence. Court system
faces severe challenge on this problem, while search
for similar criminal cases manually from historical
judgments database, is very time consuming, and is
impossible for a judge to do so for most of criminal
cases. This paper tries to use text mining technology
to provide solution.
We study the multi-criteria decision making problems
in which the information about preference values is
expressed in the form of uncertain linguistic
information. We introduce some operational laws of
uncertain linguistic variables and the concept of the
ideal point of preference values with uncertain
linguistic information, and develop an uncertain
linguistic weighted averaging (ULWA) operator. An
ideal -point-based approach to multi-criteria decision
making with uncertain linguistic information is
proposed. Finally, an illustrative numerical example is
also given to verify the developed approach and to
demonstrate its feasibility and practicality.
Among supplementary teaching methods, online
teaching is the most important one. It is not limited to
time and location, and can record the learning
behaviors, such as online frequency, online time, and
reading materials, in the learning portfolio. However,
such recordings are mostly accumulative data, which
cannot reflect the outcome of the, and provide
appropriate learning aid. The Term Frequency Inverse
Document Frequency (TFIDF) of keywords was
treated as the criterion for the quality of the posted
paper. Experimental results indicated that students
who posted more non-useful articles have poorer
learning outcome, and the quality of posted papers
and learning outcome are closely correlated.
Keywords: text mining, online teaching, subject
categorization, supplementary learning.
Data mining is often used during the knowledge
discovery process and is one of the most important
subfields in knowledge management. Data mining
aims to analyze a set of given data or information in
order to identify novel and potentially useful patterns.
These techniques such as decision trees, artificial
neural networks, rule mining and genetic algorithms
are used to discover patterns or knowledge that are
previously unknown to the system and the users. Data
mining has been used in many applications such as
marketing, engineering, medicine, crime analysis,
expert prediction, web mining, and mobile computing.
Text mining is variation on a field called data mining
that tries to find interesting patterns from large
databases. Text mining also known as intelligent text
analysis, Text data mining refers to the process of
extracting interesting and knowledge from
unstructured text.
Text mining is young interdisciplinary field which
draws on information retrieval, data mining, machine
learning, statistics and computational linguistics. As
most information is stored as text, text mining is
believed to have a high commercial potential value.
Text mining is similar to data mining, except that data
mining tools are designed to handle structured data
from databases, but text mining can work with
unstructured or semi-structured data sets such as
emails, full-text documents and HTML files etc. As a
result, text mining is a much better solution for
companies. To date, however, most research and
development efforts have centered on data mining
efforts using structured data. The problem introduced
by text mining is obvious: natural language was
developed for humans to communicate with one
another and to record information, and computers are
a long way from comprehending natural language.
Humans have the ability to distinguish and apply
linguistic patterns to text and humans can overcome
obstacles that computers cannot easily handle such as
slang, spelling variation and contextual meaning.
However, although our language capabilities allow us
to comprehend unstructured data, we lack the
computer’s ability to process text in large volumes or
at high speeds.
Start, with a collection of documents, a text mining
tool would retrieve a particular document and
preprocess it by checking format and character sets.
Then it would go through a text analysis phase,
sometimes repeating techniques until information is
extracted. The resulting information can be placed in
management information system, yielding an
abundant amount of knowledge for the user of that
system.
The problem of Knowledge Discovery from Text is to
extract explicit and implicit concepts and semantic
relations between concepts using Natural Language
Processing techniques. Its aim is to get insights into
large quantities of text data. Knowledge Discovery
from text draws on methods from statistics, machine
learning, reasoning, information extraction,

30
knowledge management, and others for its discovery
process. Knowledge Discovery from Text plays an
increasingly significant role in emerging applications,
such as Text Understanding.
Data mining tools can answer crime questions that
have traditionally been too time consuming to resolve.
They search databases for hidden patterns, finding
critical information that experts may miss because it
lies outside their expectations. The overall goal of the
data mining process is to extract information from
data set and transform it as an understandable
structure for further use.
Text summarization is helpful for trying to figure out
whether or not a lengthy document meets the user’s
need and is worth reading for further information.
With large texts, text summarization software
processes and summarizes the document in the time it
would take the user to read the first paragraph. The
key to summarization is to reduce the length and
detail of a document while retaining its main points,
overall meaning. The challenge is that, although
computers are able to identify people, places and time,
it is still difficult to teach software to analyze
semantics and to interpret meaning.
III. METHODOLOGY
To achieve the objectives of research, the steps to be
followed are as under:
Step1:-Firstly we have to collect the data which
contains the text data in the form of word and pdf
files. The major collection of the data is District Court
Ludhiana and High court Chandigarh.
Step2:-We will read the data with the help of OCR
Step3:-After this we extract the major data fields from
the digital data with the help of this tool.
Cosine similarity algorithm:-
There are many measures of similarity, shared Word
count, word count and bonus and cosine similarity.
The most obvious measure of similarity between
documents is count of their words. For an information
retrieval system, we likely to have a global dictionary,
where all potential words will be included in a
dictionary, with the exception of stop words. Every
documents used as a training data or spreadsheet will
be parsed into an entry of dictionary, showing as a
vector.
The classical information retrieval approach to
comparing documents is cosine similarity
Cosine similarity is a measure of similarity between
two vectors of n dimensions by finding the cosine of
the angle between them, often used to compare
documents in text mining. Given two vectors of
attributes with n dimensions, A and B, the cosine
similarity, -, is represented using dot(.) product and
magnitude as
Similarity=cos ( ) =A.B/||A||B||.
For text matching, the attribute vectors A and B are
usually the term frequency vectors of the documents.
The cosine similarity can be seen as a method of
normalizing document length during comparison.
The cosine similarity of two documents will range
from 0 to 1, since the term frequencies cannot be
negative. The angle between two term frequency
vectors cannot be greater than 90°. 1 meaning exactly
the same, with 0 indicating independence and in
between values indicating intermediate similarity or
dissimilarity.
IV. SELECTION OF TEXT MINING TOOL
English text (words) parsing tool have been well
developed, tools are available at popular database
management system software, while Chinese text
parsing tool are limited and still at developing status.
Chinese text is no blank to mark word boundaries as
result, indentifying meaningful words is difficult,
because of segmentation ambiguities and occurrence
of unknown words, Chinese document parsing tool.
Automatic Sentence Parsing and Semantic
Composition under E-How Net (ASPSC) in short
developed by Jiann. Researcher of the Institute of
Information Science IIS was adapted in this research.
Since the facet and evidences of criminal case are
listed in prosecutor’s indictment, to save time for
judges, can we identify? The major facet and list of
evidences of indictment document and use those to
compare with the judgments history of court, to find
similar cases and articles of law as well as the term of
punishment been applies in these cases? Use the
indictment of criminal case that send from district
prosecutor’s office, parse this document into key
words table (vector of words count that appear in
document, table1 shows the example), then use this
key word table1 (vector) to match with each vector of
court’s judgments built in trained database (i.e.
dictionary, in key word table format, table2 shows the
example).
Table 1: Determination of frequency
Table 2: Similarity between indictment and judgments
Using cosine Similarity algorithm of text mining
technology to calculate coefficient of similarity
between indictment and judgments. Base on the
highest coefficient, we will find closest judgment of
this criminal case.

31
Select highest 5 coefficients of judgments, a
referential judgment report will be generated
automatically, which provide information of similar
criminal facets and term of punishments to judges for
balance judgment. The quantity of judgments in
referential report is adjustable. Owning to indictments
of prosecutor is not open to public, hence; this
research will use police’s investigation documents
instead (similar to indictment, police’s investigation
documents have information of criminal facet,
evidences list, etc.) for training and testing key word
dictionary.
Step4:- In the last step we will have the final data
which can be used for judgment for the judge
The main idea of this paper is used for police
investigation document of criminal case send from
district police office , parse this document into
keyword table, then match this data with the trained
database in the form of key word table format of
court’s judgment. Using Cosine similarity algorithm
of text mining technology to calculate coefficient of
similarity, base on highest coefficient, we will find
the closest judgment of the criminal case.
Security and safety always rank as high priority
concerned issues in country. People always have a
request for adequate and fare punishment to those
criminal cases, while we may found most and large,
criminal case that has similar faces may received very
different punishment. This is the serious challenge to
court system. To make fare and balanced judgment,
judges must search for similar criminal cases from
historical judgments database, to find out punishments
that had been made before he makes his decision. In
year 2008 each judge of district courts in Ludhiana
and Chandigarh has 75% (average) per month. Now in
year 2015 each judge of district court in LDH have
65% (average) per month, and in High court CHD
each judge have 60% (average) per month complaint
sounds on overloading is laud and clear. Searching for
similar criminal cases manually from historical
judgments database is very time consuming and is it
not possible for a judge to do in most criminal cases
handling except some critical cases. This paper will
tries to use text mining approach to solve problems
faced by judges and people should get more aware
about the rules and punishments so that they can safe
their children from many things.
If we classify judgments of drug abuse criminal cases
into several clusters base on the similar facet of
criminal case and assign a judgment template for each
cluster base when we building up training database
for cosine similarity comparison. Then we can use
cosine similarity comparison approach to find the
nearest judgments for a target indictment that send
from district prosecutor’s office. We can then find the
best fit judgment template base on which cluster it
belongs, by the way we can digest import stuff from
indictment (feature extraction) and paste it to the
template, a draft of court judgment may generated
automatically. Judge only need to make necessary
modification instead of writing a full judgment
manually
 Constrained by time and resource, this paper will
focus on the first issue, other topic is govt. should
get more alert and make strict and more strict
rules so that nobody can think of doing any
wrong work that must be punishable.
 Collection of data regarding criminal cases.
 Making unstructured data from the structured
data.
 With the help of that making a decision
supportive and less time consuming.
V. SIMULATIONS AND EXPERIMENTAL
RESULTS
The proposed solutions have been designed using
Xilinx. The area-efficient carry select adder can also
achieve an outstanding performance in power
consumption. Power consumption can be greatly
saved in our proposed area-efficient carry select
adder because we only need one XOR gate and one
INV gate in each summation operation as well as one
AND gate and one OR gate in each carry-out
operation after logic simplification and sharing partial
circuit. Because of hardware sharing, we can also
significantly reduce the occurring chance of glitch.
Besides, the improvement of power consumption can
be more obvious as the input bit number increases.
REFERENCES
[1] Agahi H, Mohammadpour A, Vaezpour M S. Predictive
tools in data mining and k-means clustering: Universal
Inequalities. 2013; 63:3:4 Issue 3-4;779-803.
[2] Bhushan J , Pushkar W, Shivaji K, Nikhil K .Searching
Research Papers Using Clustering and Text Mining‖
International Journal of Emerging Technology and
Advanced Engineering Website: www.ijetae.com, 2014;
4:4.
[3] Chauhan SR, Desai A. A Review on Knowledge
Discovery using Text Classification Techniques in Text
Mining”. 2015; 111:6.
[4] Ghalib M, Vohra S, Vohra S, Juneja A. MINING ON
CAR DATABASE EMPLOYING LEARNING AND
CLUSTERING ALGORITHMS. International Journal of
Engineering and Technology (IJET).
[5] Garrette D, Mielens J, Baldridge J. Real-World Semi-
Supervised Learning of POS-Taggers for Low-Resource
Languages. Proceedings of the 51st Annual Meeting of the
Association for Computational Linguistics (ACL-2013).
583-592.
[6] Funmilayo K, Afolashade K, Oludele A. Text Mining
Approach in Curtailing Cyber-Crimes in Nigeria.
International Journal of Computer Science Trends and
Technology (IJCST), 2015; 3:6.
[7] Ghosh S, Dubey S K. Comparative Analysis of K-Means
and Fuzzy C-Means Algorithms. International Journal of
Advanced Computer Science and Applications, 2013; 4:4.


Text mining on criminal documents

More Related Content

What's hot (19)

Similar to Text mining on criminal documents (20)

Recently uploaded (20)

Text mining on criminal documents