Ontology Based SMS Controller for Smart Phones

Text analysis includes lexical analysis of the text and has been widely studied and used in diverse applications. In the last decade, researchers have proposed many efficient solutions to analyze / classify large text dataset, however, analysis / classification of short text is still a challenge because 1) the data is very sparse 2) It contains noise words and 3) It is difficult to understand the syntactical structure of the text. Short Messaging Service (SMS) is a text messaging service for mobile/smart phone and this service is frequently used by all mobile users. Because of the popularity of SMS service, marketing companies nowadays are also using this service for direct marketing also known as SMS marketing.In this paper, we have proposed Ontology based SMS Controller which analyze the text message and classify it using ontology aslegitimate or spam. The proposed system has been tested on different scenarios and experimental results shows that the proposed solution is effective both in terms of efficiency and time.


INTRODUCTION
Mobile phones were initially developed to make and receive calls while being mobile using radio link. Later on, services like text messaging (SMS), multimedia messaging (MMS) were added in the mobile phone devices. In the last two decades, mobile phones have evolved and have become smarter / intelligent devices commonly known as smart phones [1,2]. Smart phones are built on mobile computing platform and usually have advanced computing abilities / connectivity as compared to the simple mobile phones [8]. Initially smart phones were developed with the integration of mobile phone and personal digital assistant (PDA) functions. These smart phones include number of exciting features like touch screen, mobile web browser (i.e. access websites on mobile phone) and WiFi (i.e. access internet using wireless connection) support.
In year 2000, high resolution touch screen smart phone named Ericsson R380 was released which has its own Operating System. This was first ever smart phone with its own OS, the Operating System used was Symbian OS. In 2005, Google entered into the mobile market with the help of an open source operating system for smart phones called Android. In 2007, Apple [9] introduced a smart phone named iPhone [10] which made big change in the history of smart phones development; Apple development their own Mobile Operating System named as IOS for iPhone and this OS is not open source. Therefore, Android operating system is supported by most of the smart phones companies (such as HTC, Samsung, Sony Ericson).
Along with the launch of iPhone, Apple introduced AppStore (Application Store) where 3 rd party applications were hosted for distribution (i.e. single platform distribution). Before, Apple AppStore, smart phone applications distribution were largely dependent on third-party sources that developed the application(s) such as GetJar, Handmark, Handango, PocketGear, etc. Application development for Android OS is greatly increasing as compared to IOS because 1) development toolkit is free, 2) Android is open-source, therefore it's easy to integrate applications and 3) Android software suite allows easy integration with Google applications such as Maps, Calendar, web browser etc. Android based smart phones are giving great competition to iPhone.
Text analysis includes lexical analysis of the text and has been widely studied and used in diverse applications. In the last decade, researchers have proposed many efficient solutions to analyze / classify large text dataset, however, analysis / classification of short text is still a challenge because 1) the data is very sparse 2) It contains noise words and 3) It is difficult to understand the syntactical structure of the text [21,22,25,28]. The concept of Short Messaging Service (SMS) was developed in the Franco-German GSM cooperation in 1984 by Bernard Ghillebaert and Friedhelm Hillebrand [11]. SMS is a text messaging service on the phone, web or mobile system and mostly used data application is SMS text messaging. SMS nowadays is also used for direct marketing also known as SMS marketing.
In the last few years, many SMS managers have been developed for managing the SMS on smart phones and the most of them focuses on Spam filtering, Scheduled SMS and automatic-Reply generation. Few popular android applications are 1) Anti SMS Spam: It is a spam filtering application and spams all incoming SMS from unknown numbers when Spam filtering is turned on. 2) Schedule SMS: It is scheduled SMS application and gives time, date, recipient number and text (SMS content) option to the user. The application sends the SMS to the recipient on specified time and date specified by the user. 3) SMS Auto Reply: It is an Auto Reply SMS application which sends an automated reply to all the incoming texts when auto reply is turned on. The content (text) of auto reply is selected / configured by the user.
Ontologies have been widely used for knowledge representation / sharing and have been used in diverse areas [23,24,26,27]. Ontology based SMS Controller is an Android Based Application developed on Android Jelly Beans 4.1, the proposed solution is all in one SMS manager and includes some previous features with advancements as well as some www.ijacsa.thesai.org new and exciting Features like ontology based SMS spam detection, Group chat etc. The default android messaging application gives few options to user such as send message / receive message / save message etc. whereas the proposed application provides some additional features in addition the default features. The major features of the proposed application are:  Automated text replies to messages when a profile is activated  Scheduled SMS sent on specific dates and events  Group chat including multiple users like we do in different messengers  Content based Spam filtering The above features make the proposed application unique as these features are missing in the existing applications. The remainder of this paper is organized as follows. In Section 2, we present brief overview of related work, this section is followed by the discussion of the Ontology based SMS Controller architecture including the SMS text analysis and classification method. In Section 4, the simulation and experimental analysis of proposed solution is presented. Finally, the conclusion is drawn in Section 5.

II. RELATED WORK
With the evolution in Smartphone era, leading IT companies and researchers have proposed many efficient applications for the same. In this section, we will review few related applications developed for managing SMS on android platform. [12] is an android based anti-spam application with a private box. Its spam feature helps filter unwanted messages from any sender. [13] is an android based SMS scheduling application with the following key features: a) It helps schedule SMS/MMS messages at specific times or at regular intervals e.g. daily and b) It supports blacklist (i.e. deletes incoming SMS / MMS from number in blacklist which also helps block spam messages.

Key features of this application are: a) Can block SMS from unknown numbers, and b) User can create a block list and can add existing contact or new numbers in the block list. 2) Handcent SMS by Handcent Market
3) SMS  [14] is an auto reply application which sends reply to each incoming SMS when the auto reply option is enabled. The user has to configure the reply to be send when the auto reply option mode is turn on.

5) Intelligent auto reply by John Tsau [15] is a rule based SMS application that has Auto Reply and Auto Forwarding features. It automatically replies to the SMS and missed calls according to the rules set by the user. 6) GO SMS PRO by GO Dev Team [16] supports the features of scheduled SMS and Group Texting.
The previous applications include most of the exciting features but they have the following limitations:  What if a user does not want to spam all the SMS from unknown numbers?
 What if a user wants to spam the SMS from some unusual numbers only?
 What if a user wants his application to auto reply to a certain Group?
 What if a user wants to send different replies to different group of recipients?
 What if a user wants to have all these features in one application?
Ontology based SMS Controller has Solutions to all these questions. It auto reply to a certain Group and sends different replies to different group of recipients. It gives solution for detecting spam SMS using content analysis. Above all these features are all integrated in one application so that a user can easily manage all the features from one application. Plus it includes new features like Group chat and Auto scheduled SMS by synchronizing the events in the Calendar. The key features of the proposed application include:  Pre-processing: In the first step, for each incoming SMS, the Ontology based SMS Controller validates the sender number with the spam blacklist numbers, if the number is found in the blacklist numbers, the SMS is send to SPAM folder without further processing. The Ontology based SMS Controller also provides user the option to SPAM all SMS messages from unknown numbers or specific numbers or weird numbers. If this option is selected by the user, all SMS belonging to these categories will be send to SPAM folder without further processing. In Step 2, all standard stop-list / stemmer words like ("is", "the", "on", "and", "in", "with", "for", "by"…) are eliminated from the SMS Text. In Step 3, homogeneous words like {("chat", "chatting", "chatted"), ("Advertize", "Advertizing", "Advertized")} are all substituted by the single word "chat" and "Advertize" respectively. Also, multiple entries for each word are eliminated from the SMS text.

 Content Analysis: This module uses the filtered SMS
text from the previous step which contains n keywords where each keyword can express n possible meanings.
In order to assign proper meaning to each keyword, every keyword is compared with every other keyword and most related sense (i.e. semantically related) is selected. To calculate the most related sense the shortest path (i.e. minimum number of nodes present in the path connecting the keywords is used); WordNet [18] is used for this purpose. In the next step, Concept set is generated which contains either the original keywords or Lowest Super Ordinate (LSO) for each pair of keywords, the selection depends on the parameter h, for more details please see [19].

B. Group Chat
The prerequisite of using this feature is that all participating users should have Ontology based SMS Controller installed on the smart phone. One of the application user has to start the Group chat by sending "join group chat" invitation to the others. Invitation is sent through SMS message and for this purpose a special SMS is send to the invitee which Ontology based SMS Controller interprets and asks the user to join the group chat. The user can accept or reject the request, if the user accepts, the details of new user is send to all the active members of the group chat and chat window is loaded on the new user's Smartphone.
Similarly, if any active user during the group chat closes the application, the details of disconnected user is send to all members of the group chat. Each group chat is assigned unique chat code to the same and each SMS message send / received from chat window contains this unique chat code, which makes it easy to identify; to which chat this message belongs. Each member sets a nick at the start of chat, and these are displayed on the chat window instead of the numbers.
Each incoming SMS is interrupted by Ontology based SMS Controller and it validates the type of the SMS message (i.e. Invitation, Chat Message, or Normal Message) and perform actions accordingly. If the SMS is chat message, it forwards the same to the corresponding chat window and delete it from the inbox. If the SMS is invitation SMS, it displays the invitation to the user and wait for the response. Based on user response it either opens the chat windows or sends rejection message. When a user joins or leaves the chat, all other members are informed and the list of chat members is updated accordingly.

C. Auto Reply
Ontology based SMS Controller sends auto-reply according to the user-defined profiles; the user is responsible to create auto-reply groups and reply messages for each group. If the auto-reply mode is on and no auto-reply profile is activated, a default auto-reply message is sent otherwise the user defined auto-reply message according to the profile is send. User can create groups and add numbers in these groups from contact list.

D. Event-Based Messages
Ontology based SMS Controller automatically synchronizes itself with the calendar and generates automatic SMS based on Events. User is responsible to define event(s) by setting date / time of the event(s) and the message to be send. User can add group(s) to an event by selecting from list of available groups. Message defined against the event is send to all members of the group(s) associated with the event. Birthday event is predefined in the application, which picks the birthdays of the contacts (if available) and create birthday event for each of them. User can define the birthday message for the birthday event, if no message is defined; default birthday wish is send automatically on respective birthdays. Initially for the experimentation, we built the ontology concepts using one hundred known spam messages; afterward we tested the proposed solution on large number of SMS, figure 5 shows spam detection percentage over number of SMS, as shown in figure 5 the proposed solution spam www.ijacsa.thesai.org detection percentage over number of SMS increases as ontology knowledgebase is enhanced (i.e. new spam concepts are updated in the ontology).       figure 9(b) is shown on the sender device.After handshaking the group chat is started and chat window appears on each participating device (i.e. Sender & Receivers) as shown in figure 10.User can write the text in text Box and use the send button to send the message. The message is send to all recipients by using normal SMS and is shown on each participating device. New members can be added at any time using the invite more feature.  The user can create new groups, just he needs to give the title of the group and add contacts in the group as shown in figure 11(a) and (b) respectively. Contacts can be added to group by going to the desired group and selecting "add contact" from menu. Similarly user can create new profile(s), each profile contains profile title and auto-reply SMS. The user needs to enter the title of the profile and auto-reply SMS as shown in figure 12(a) and (b) respectively. Auto-Reply SMS can be added by going to the desired profile and selecting "Select auto-reply SMS" from menu. Auto-reply mode can be activated by clicking the toggle (on/off) button in figure 13. Left window of figure 13 shows the view where auto-reply mode is off, no group and profile selected. Right window of figure 13 is showing the view where mode is on, one group is selected and profile is activated.
Spam Filter can be turned on by clicking toggle button on the spam window. Spam folder contains all the spam messages. Spam List contains all the numbers which are marked as spam. Spam Settings allows user to spam messages from weird numbers or unknown numbers by checking the checkboxes. A group can be added and marked as spam as well as shown in figure 14.
Similarly new events can be added by the user specifying event title, date / time, profile / SMS and group(s) / numbers associated with the event. The application sends the specified SMS to all the members associated with the event on the event date / time.

V. CONCLUSION
With the evolution in Smartphone era, leading IT companies and researchers have proposed many efficient applications; one of them is SMS Manager which helps to manage the SMS on smart phones. In this paper, we proposed Ontology based SMS Controller which is all in one SMS manager and includes new features like content based SMS Detection, Group chat etc. SMS Spam classification algorithm of Ontology based SMS Controller analyses the text of SMS and uses ontology to classify it as Spam or legitimate. The proposed algorithm has been tested on large number of test cases; the experimental results are satisfactory and supports the implementation of the solution.