Abstract—These days, social media has played a significantrole in daily life of all people and ages in order tocommunicate as well as express their thoughts and feelings.In this paper, the authors have studied user data from socialmedia (Facebook) whose shared posts are positive, and alsothe negative side posts that may lead to negative affectpersonally or can be further extended to the community andnation level. The purposes are to identify users who havecommented on the negative side that may be a lawbreaker onComputer related crime. On this which beneficial aboutinvestigation for legal proceeding and it facilitate for thepolice or people who take a part in the operation on law. Italso contributes in the community at large to peacefulness.
The effective Naïve-Bayes classifier is used in order toclassify these two user groups. It significantly shows thatanalyzing social media data by using Naïve Bayes modelpresented sharing positive and negative views accurately aswell as reflects satisfied results.Keywords-Forensic Analysis · Social Media · Facebook ·Naïve BayesI.I NTRODUCTIONDigital development in modern society and the use ofInformation Technology which are very useful andcreative become a big part in our daily life.
On the otherhand, it may lead to negative affect when it is usedimproperly in person or can be further extended to thecommunity and nation level where there is a very highpossibility of the transaction.In recent years, the social media has significant role toeveryday life of all people and ages to communicate aswell as sharing their thoughts and feelings. ForensicAnalysis is the application of scientific principles andtechniques to the legal process in order to investigate,testify evidence, and lead offenders to criminal justice.Without forensic evidence, many offenders in complexcriminal cases are still free and may repeat offendingagainst the laws and harm the others.Thus, in many developed countries such as Japan,European countries and the US have applied existingscientific knowledge and technology to identify evidencesin criminal trials to gain true scientific results which isvery useful in criminal investigations.
Particularly inJapan, shows it having over 90% of the murderers arearrested by applying scientific equipment and invented andinnovated technology to be used in forensic science. Infact, if we bring information technology and socialnetwork process intotransaction management in database, and then apply tothe forensic science, it will lead greatly to the benefits evenin forensic process, fairness human right, as well as orderin society.Recently, many researches present the social mediadata that are analyzed in or-der to apply for correlationanalysis and also the further result of existing data.Particularly from the study of researchers in forensic field,data analysis in forensic 1-8 shows that the social mediadata can be divided into major groups to be used in thestudy such as the researches that analyze data in text form,and the researches that analyze data in image form. InThailand todays, the use of social media is un-closed thatthe information of social media users in text form ispopular and it also reflects the use of social network suchas Facebook is highly accepted and its growth rate ismultiple.Therefore, the authors propose to use data from socialnetwork (Facebook) applied with the field of forensic,analyzing with Naïve-Bayes classification algorithms inorder to clarify data into two specific groups.
The firstgroup relates to sharing their thoughts and feelings on thepositive side, and the second one relates to negativeposting. The purposes are to identify users who havecommented on the negative side which may causeunwanted effect personally or can be further extended tothe community and nation levelLITERATURE REVIEWUraz Yavanoglu, Busra Caglar, Ozlem Milletsever,Medine Colak, Semra Cakir and Seref Sag_roglu proposedthe Intelligent Approach for Identifying Political Viewsover Social Networks. It is a research-based analysis ofpolitical views by analyzing Social Network data throughArtificial Neural Networks: ANN model and Data mining.The data used in this research is taken from Twitter whichis a public data. Therefore, this work helps to analysisthoughts and ideas from Twitter users both supporting oropposing the government 1.Chris Howden, Lu Liu, Zhijun Ding, Yongzhao Zhanand KP Lam proposed the Moments in Time: A ForensicView of Twitter which the Twitter is carried through thePython IDLE client with MySQL database and display thedata from Twitter via the SQL Statement 2.Shankar Setty, Rajendra Jadit, Sabya Shaikh, ChandanMattikalli and Vma Mudenagudi proposed theClassification of Facebook News Feeds and SentimentAnalysis which presented a system for classification ofFacebook news feeds andA learning based classifier is built using variousclassification algorithms such as Bi-nary LogisticRegression, Naive Bayes, Support Vector Machine(SVM), Bayes Net and J48 and This experiments on thelive news feeds showed that the proposed approach couldachieve significantly improved performance for structuringthe data on Facebook using SVM classifier learning model3.