144 Chapter 7 Social Media and Social Networking: The Present and Future Directions Ahmed Elazab Cairo University, Egypt Mahmood A. Mahmood Cairo University, Egypt Hesham Ahmed Hefny Cairo University, Egypt ! ABSTRACT i Social media is a powerful communication tool that facilitates the interaction and provide an efficient interconnection among different roles in many fields such as business and media. The power of social media forced its responsibility for the vast dissemination of different information during real time events. Many social networks have emerged since the 905; however, many of these networks have been abandon while the success of others in providing intelligent and active communication made them the most famous recently. Some examples of these successful social networks are Facebook, and Twitter. In this research, we provide the readers with the main concepts of social media and social networks, and their relation with other fields We also discuss the current situation with providing the emerging trends and challenges of both fields. DOK: 10,4018/978-1-5225-2897-5.chO07 Cony ©2011 Go. Coping dg ri cn em tow is peli GL Fm |social Meda and Social Networking INTRODUCTION The internet has changed the way of communication forever, the way people buy things, work, and even socialize, Although some are convinced that internet decrease the communication level among people, however, others think totally the opposite in the presence of many electronic communication methods suchas social media, Social media playsa vital role in all online aspects now, including personal communication, business, economic, it even affects political aspects seriously. In Canada, a study considered the impact of social media has found a raise of the economic level from 1.9% in 2011 to be 3.8% in 2012 (Industry Canada, 2012). In business field, social ‘media effect is increasing tremendously, many businesses now are getting benefits from people's opinions to raise the level of their products and services, which is considered one of basis in their growth. For example, the Public Health Agency of Canada” has been using social media to provide valuable information considering the public health with demonstrating the threats and suggesting health plans. The idea of Social Media is top of the plan for some business officials today. Chiefs, and in addition advisors, attempt to recognize routes in which firms can make beneficial utilization of uses, for example, Wikipedia, YouTube, Facebook, Second Life, and Twitter. However in spite of this enthusiasm, there is by all accounts exceptionally restricted comprehension of what the term “Social Media" precisely implies; this article plans to give some elucidation. Kaplan and Haenlein (2010), start by depicting the idea of Social Media, and talk about how it varies from related ideas, for example, Web 2.0 and User Created Content, In light of this definition, Kaplan and Haenlein then give a characterization of Social Media which bunches applications as of now subsumed under the summed up term into more particular classes by trademark: collective tasks, online journals, content groups, interpersonal interaction locales, virtual diversion universes, and virtual social universes. At long last, they display 10 recommendations for organizations which choose to use Social Media (Kaplan & Haenlein, 2010) With the rapid evolution of content and communication styles in social media, this emerging media has become a powerful communication channel, as evidenced by many recent events like “Egyptian Revolution” and the “Tohoku earthquake and ‘sunami, Textis changing too. Different from traditional textual data, the textin social media is not independent and identically distributed data anymore, A comment or Post may reflect the user’s interest, and a user is connected and influenced by his friends. Based on internet users’ feelings that were measured from their texts, (Hu & Liu, 2012) was able to investigate their political opinion as well as their confidence. 145‘Social Media and Soca Netverting Kuan-Yu Lin applies arrange externalities and inspiration hypothesis to disclose why individuals keep on joining SNS. This review utilized an online surveytoleadexac, research, and gathered and investigated information of 402 specimens by auxili condition displaying (SEM) approach. The discoveries demonstrate that delight ig ‘the most compelling component in individuals’ proceeded with utilization of SNs, trailed by number of associates, and convenience. The quantity of companions and, saw complementarity have more grounded impact than the quantity of individuals on saw benefits (convenience and happiness). This work additionally ran grouping ‘examination by sexual orientation, which discovered striking contrast in both number of associates and number of individuals amongst men and ladies. The quantity of ‘companions is an imperative component influencing the proceeded with goal to use for ladies yet not for men; the quantity of individuals has no noteworthy impact on delight for men. The discoveries recommend that sexual orientation contrast likewise delivers diverse impacts. The ramifications of research and exchanges gives reference to SNS administrators in advertising and operation (Lin & Lu, 2011). However, sharing this amount of personal information may have its consequences, as a person may be involved in a situation as a result of posting inappropriate post, or he may receive incorrect information about himself which is posted from others. For example, Facebook gives the user the ability to restrict friends or set page viewing, however, the subject of privacy still an issue in social media. Other problems arise in social media, claiming to be another person is one of the main issues, a person may create a Facebook page claiming to be someone else, using bis photos and other known personal information which may cause a lot of problems to the original person. To conclude, we can claim that privacy is lost in social medi ceven with the current status. ‘Social media web sites contain various types of services and thus create different formats of data, including text, image, video etc. For example, the media sharing sites Flickr and YouTube allow to observe what “ordinary” users do when give the ability to more readily incorporate images and video in their everyday activity People engaged in the creation and sharing of their personal photography. AS # result, a large amount of image and video data is archived in the sites. Besides, blogging sites, the users post frequently and create a huge number of textual / text based data; in social bookmarking sites, users share with each other tags and URLS ‘The chapter presents the basic understanding of these analysis approaches with demonstrating different work performed by researchers on the aspect with the required enhancements for further development. This structure of the chaptet is as follows: presents basic definitions for social media and social networking: ® comparison between them. Also it provides the relation between the information extraction and social media, and then presents different information extraction approaches with focusing on the approaches applied in social media by presentiN 146‘social Medla and Social Networking different work done in this research area. And the text analytics direction in social media is discussed with presenting the issues arise in this field in section 7, and finally the conclusion is discussed. ‘The main focus of this chapteristo provide an understanding of the basic concepts, of social media, The chapter provides an understanding for different aspects related to social media, with presenting different directions for social media analysis. BACKGROUND Social media and social Networking are two paradigms that usually used for one another by users who do not actually understand the scientific concept that differentiate the two terms. In this section, we demonstrate the basic concepts of both social ‘media and social networking, Social Media Social media is nota new concept, ithas been evolving since the beginning of human interaction where the social media has transformed the interaction and communication of individuals throughout the world. However Hu and Liu (2012) defines social media as applying a set of online tools to support the efficient communication among users for sharing information and experiences (Hu & Liu, 2012). The social media sites play a very important role in current web applications. Social media such as blogs, microblogs, discussion forums and multimedia sharing sites are increasingly used for users to communicate breaking news, participate in events, and connect to each ‘ther anytime, from anywhere. Benefits of Social Media OSN acts as a powerful way to communicate also facilitate open forms of communication. Social media is best for many situations, the following are some examples: On the level of employees, Social media enables employees to shate project ideas and work in teams effectively, which helps in sharing knowledge and experiences, it also promotes open communication between employees and management. Also, itencourages supporting members, or part of the company’s employees, to become members of a well-recognized community (Edosomwan, 2011). On the level marketing, social media becomes a good place for discussions and 4 classic goal of marketing and communications, however, also OSN promotes better content, such as webcast and videos, than just simple text. OSN also helps 147‘Social Media and Social Networking ‘0 communicate collaboratively between current and new customers, in receiving feedback, product definition, product development, or any forms of customer service and support Social Media Platforms Nowadays with the evolution of technology, supported by global and speedy communication network, online social media has been growing rapidly in the form, of collaboratively created content whichis presents new opportunities andchallenges to both producers and consumers of information, With the large data presented by various social media services, text analytics provides an effective way to meet users! diverse information needs. There are many types of social media prevented based on their category as shown on Table 1 (Treem & Leonardi, 2012) In the Sensis Social media report (Social, 2015), and Social media update report Duggan and collegues (2015) which are conducted in 2015, adiscussion for some of the most popular social networks is presented on Australia and American societies respectively. The first report showed that 93% of Australians access Facebook, 28% of them access LinkedIn, 26% for Instagram, 17% for Pinterest, and 17% for ‘Twitter on the daily basis. While Americans have different statistics as 71% of American access Facebook, 28% of them access LinkedIn, 28% for Instagram, 26% for Pinterest, and 23% for Twitter on the daily basis. This comparison is presented graphically in Figure 1. Table 1. Social Media Platforms (Treem & Leonardi, 2012) lpn heNn Pe patos erengea cn Mourne [se — ea | oe al feos Uh sce 148‘soci! Media and Social Networking Figure 1. Comparison Among Social Media Plasform Moreover, in the American adult report (Duggan, et al.,2015), the study revealed that 42% of adult Americans usually use more than one social media sites in 2014, while this percentage increased in 2015 to be 52%. While for Australians, a different comparison is presented in the Sensis social media report (Social, 2015)as it is ‘mentioned, thatthe percentage of the Australians who check social media daily was 41% in2014 which increased in 2015 to be 45%. Senisi report (Social, 2015)focused ‘on business field, it applied a study on 1,100 of businesses and 800 consumers, and this study revealed that social network is currently one of the essential methods that businesses depend on. Finally as shown in the Figure 1, it is clear that Facebook still has the lead for popularity in general in both report segments Social Networking In the 1990s, many social networking sites appeared. The first well-known OSN site, is called “Six Degrees”, Other examples include “BlackPlanet, Asian Avenue, and MoveOn”. These are, or have been, online socialsites where people can interact, including sites for public policy support and a social network based on a web of contacts model. In addition, blogging services such as “Blogger and Epinions” ‘Were created. “Epinions” is a site where opinions about products are submitted by different customers, these opinions support other consumers to have a near correct decision (Golbeck, 2005). In the 90s, the software applications named “Third Voice and Napster” (Edosomwan, 2011) were created but have been removed from the market. Third Voice was a free plug-in to allowed users which provided them with ‘an access to post comments onWebPages. Napster was a software application that 149‘Soci! Media and Social Networking allowed peer-to-peer file sharing in which users were allowed to share musi files bypassing normal distribution methods. However, in the end, this situation was determined to been violation of copyright laws (Edosomwan, 2011). However, many social networks such as Yahoo 360, YouTube, and Facebook are introduced and are currently popularly used. Facebook was initiated only for the Harvard community, however, it became accessible by high school students in the year of 2005, then later in the 2007 the rapidly growth of Facebook was gained over | million new users every week (Ellison, 2007) Other method for interaction, among people by “Yahoo! 360” which was a website launched by Yahoo! Inc, in ‘which people could create a profile with photo albums and interact with other people ‘with similar interests or get in contact with some of their friends like in any other social network. Another idea is “YouTube” which is provided and has currently a) wide usage. The idea started when some of PayPal employees wanted to upload their videos, then they created the YouTube system (Rautio, 2012). Social Media vs. Social Networking In order to discriminate between the two terms, in this section, we will provide al short comparison between Social Media and Social Networks according to some! determined criteria, As a start, Social Media can be defined as a strategy and an| access forbroadcasting, while Social Networking is a tool and a utility for connecting) with others. The differences between the two terms are not only semantics but also} Table 2. Comparison between Social Media and Social Networking =a — ————rerecig ————) [eter Aa vn ay Tw | A aT cc a FRE SE rma wih elses. | en nets eine pert ‘iam om 213), ‘rrp Brug commy ies (Cote 30% ls, 10, ‘Soro wea he ca ‘how whch. “lion we \ ‘Src det Sn Dp [EGE = Rea peat maton ery eae oe ao ‘voces |e Kt me men | Ne sy es aso | ot iyo the ty |e sacl mest they sae | | Sree ey fee | |Spas seers cease __ aS al 150social Media and Social Networking in the features and functions which is included in these websites by their creators, these features dictate the way they are designed and used. Moreover, the differences between social media and social networks are summarized as shown in Table 2. MAIN FOCUS OF THE CHAPTER Information Extraction From Social Media Ingeneral information extractionis.afieldthatis concemed with obtaining information from different sources. Focusing on online sources such as online databases and web resources services including websites, according of the dynamic nature of the World Wide Web, it became important to find tools for information extraction from the web as end users and application programs have some difficulties when it comes to finding useful information (Alim, et al., 2011), Extracting information from social mediahad applied different learning approaches as HTML is the common language for implementing Web pages and it is widely supported by The World Wide Web Consortium (W3C). HTML pages can be as a formof semi-structured data in which information follows anested structure (Ferrara, tal.,2014). This section present different research that have been presented to apply «different information extraction approaches on social media. Generally the evolution of technology, supported by global and speedy communication network online Web services, has been growing rapidly which presents new opportunities and challenges to both producers and consumers of information. The volumes of click stream and client information, gathered by web- based associations in their everyday operations, have come to galactic extents, Analysis of such information helps these organizations to focus on the life-time estimation of customers (Mobasher, 2006). Data mining technique that automatically discovers or extracts the information from web documents in general usually consists of following tasks (Srivastava, et 4, 2005); (Bhisikar & Sahu, 2013). * Resource Finding: It involves the task of retrieving intended web documents. Itis the process by which we extract the data either online or offine resources available on web. Information Selection and Pre-Processing: It involves the automatic Selection and preprocessing of specific information from retrieved Web resources, This process transforms the original retrieved data into information. The data is transformed into useful information by using suitable ‘tansformation. The transformation could be renewal of stop words, or it may 181Social Media and Social Networking be aimed for obtaining the desired representation such as finding particular format of data. ‘© Generalization: It automatically discovers general patterns at individual web sites as well as across multiple sites. Data Mining techniques and machine learning are used in generalization. © Analysis: It involves the validation and interpretation of the mined patterns, It plays an important role in pattern mining. A human plays an important role in information on knowledge discovery process on web. However, in the extraction process, many issues arise for the target of discovering useful information from online pages. One of these issues considering data Tepresentation, as website pages can be found in different formats. HTML is designed for present unstructured information, while XML and XHTML are intended for more organized information which elements help the parsers of web crawlers to communicate withthe site pages’ substance all the more proficiently (Alim, et l., 2011). Information Extraction Approaches Researchers began to use statistical techniques and machine learning algorithms to automatically create information extraction systems for new domains. In the following subsections, overviews of information extraction approaches are presented. In general, approaches of information extraction can be under the category of either supervised, weakly supervised, or unsupervised learning approaches. Supervised Tearning can be applied for extracting different information including patterns, rules, and sequential information. Moreover, weakly supervised and unsupervised learning ‘methods for information extraction are applied for more global or discourse-oriented ‘approaches to information extraction. The following subsections will demonstrate the work performed by researchers in this field. ‘Supervised Learning of Extraction Patterns and Rules ‘Soderland (1999) developed an approach to apply learned rules on text which had been annotated with the information to be extracted. However, the annotated text required for training is often difficult and time consuming to obtain. An alternative approach is to use weakly supervised learning algorithms, these do not require large amounts of annotated training data and rely on a small set ‘of examples instead. These approaches greatly reduced the burden on the application developer by alleviating the knowledge acquisition bottleneck. Weakly ‘supervised algorithms have the 152Social Media and Social Networking te Ch saving only small amounts of annotated training data (Stevenson & Several systemsuserrule learning algorithms to automatically generate information extraction patterns from explained text corpora, Relational learning methods have also been used to learn rule like structure for information extraction and applied by more researchers such as (Roth, Yih2001 Cali & Mooney 2003; Bunescu & Mooney, 2004; Bunescu & Mooney, 2007) On the same way Mooney and Cali used relational learning methods to generate information extraction rules, where each rulehasa pre-fillerand post-fillercomponent each component is a pattern that consists of words, POS tags, and semantic classes (Mooney, 1999). Another research by (Dahab, et al., 2010) proposed an extraction method for extracting Arabic patterns representing concepts. The research aimed to extract the concept in the agriculture domain with avoiding using the required tools for information extraction such as stemmer and lexicon, However, the proposed method can be further enhanced to extend its output to have a complete hierarchical tree or extract semantic relations, A recent research by (Hassan, et al., 2014) in which an approach was proposed to propose an answer to user's queries over the web. The approach combined three information extraction approaches to improve the efficiency and accuracy of the extracted answer. These approaches are “Rhetorical Structure Theory (RST), the query processing approach, and the Network Representation approach (NRA). (Hassan, et al., 2014) ". However, the proposed approach raised the problem of the synonyms and anatomies in the text to focus for finding a solution. Moreover, another research by the same authors in 2015 (Hassan, et al., 2015) in which they conducted a classification method for Arabic document for the same answering user query target. Their aim was to classify the document to avoid redundancy in the extracted information for answering the query, however, the research also raised the problem of the lack of resources such as the corpora lassifier Models Supervised Learning of Sequential ‘There are other approaches that view information extraction as a classification Problemthatcanbe tackled using sequential learning models. Instead of using explicit, Patterns or rules to extract information, a machine learning classifier is trained to Sequentially scan text from left to right and label each word as an extraction or a on-extraetion, Atypical labeling scheme is called “IOB” is presented by (Hobbs & Riloff, 2010) in which each word is classed as an ‘T’ if it is inside a desired extraction, ‘O” if itis 153Social Media and Soci Networking outside a desired extraction, of ‘B”ifitis the beginning of desred extraction The proposed scheme has been applied for extracting facts about pee ing incident, ‘Also Gentile, Lanfranchi, Mazumdar and Carvegnain 1) presented an approach to automatically and dynamically provide a model for user expertise from informal communication exchanges. The research depended on generating semantic user profiles from emails (and more generally from any textual user generated content) guaranteeing flexibility, dynamicity and providing ways to connect these data with Linked Open Data (LOD). Cui (2014) introduced different techniques such as Natural Language Processing to extract traffic information from text based on data hidden in social media. Some examples of the extracted information is which social media available platform in the mobile Internet (weibo.com) which is a Twitter equivalent in China, The research developed a prototype system that published and captured traffic status through an Android based application. On widely view, Gattani and his colleagues in (2013) described an end-to-end system for industrial that performed element extraction, linking, classification, and tagging for social data. By generate and use contexts and social signals to improve task accuracy, and how the system scale the entire Twitter. Their experiments showed that the system outperforms current proposed approaches such as a Wikipedia- based global “real-time” knowledge base that is well suited for social data). They showed that while not perfect, the system has proved useful in a variety of real-world applications. They presented thatitis important to exploit contexts and social signals to maximize the accuracy of such systems. Hamasaki and collegues (2006) related an integrated method for social network extraction including three kinds of networks, they are: user-registered Know link network, Web-mined Web link network, and face-to-face Touch link network, however, their integration needs improvement of network integration and advanced applications. Also Arjan and his colleagues in (Alim, et al, 2011) developed an automated ‘web crawler using the ruby programming language. The crawler would visit profile pages based on a randomly generated list of id numbers using the rand function of Microsoft excel. Regular expressions were used to collect the relevant bits of dat for classify age differences in online social networking Weakly Supervised and Unsupervised Approaches While supervised learning techniq an information extraction syste texts still requires a substantial extraction of information can be lues reduced the manual effort required to initiate m for a new domain, However, annotating training investment of time, and annotating documents fot fake complex. Also if we have a set of documents 8 154‘Social Media and Social Networking atraining setto the system. The documents are either relevant (contain the description of an event relevant to the scenario) or irrelevant, However, the documents are not annotated and the algorithm does not have access to this information (Srivastava, et al,, 2005) since 1980 the attempt research to extract data from the Web are two of strategies emerged learning techniques and knowledge engineering techniques also called learning-based and rule-based approaches, respectively. These approaches depends on domain expertise itneed programming experience anda good knowledge of the domain in which the data extraction system (Chang, et, al. 2006). Statistical and Learning Methods Learning in natural language processing, statistical methods have gained new popularity, and are being applied to new domains. They are usually characterized by using large text corpora and performing some analysis which uses primarily the text characteristics without adding significant linguistic or world knowledge. An example of the current widely used text corpora which have been built is Brown corpus. Annotation of corpora with part of speech tags or parse trees has been @ focus of corpus based language analysis. Additional important application areas of statistical techniques to written natural language are thesaurus-building (or lexical clustering) and probabilistic grammar learning. Statistical techniques that have been used for these tasks are n-gram techniques, unsupervised clustering and hidden Markov models (e.g. a special case is grammar induction, which uses context-free ‘grammars in addition to probabilistic information from text (Hobbs & Riloff, 2010) Rajiv, Pfeil, and Zaphiris (2008) developed an automated web crawler using the Ruby programming language. The crawler depended on the visits of the profile pages based on a randomly generated lis of id numbers using the RAND function of Microsoft excel. Regular expressions were used to collect the relevant bits of data. Challenges of Web Data Extraction Techniques Different approaches are applied for information extraction from web such as natural language processing, machine learning, and logic. Many factors affect selecting the suitable web extraction method, a major factor is the dependency level of the application domain features, and consequently it then affects the proposing of an effective solution, Some of the key challenges can be addressed for web data extraction systems. Generally, aneed for experts is required for performing the extraction task, therefore, one of the main challenges is applying automatic techniques for extraction, although ne of the main targets is to minimize the human interference, however, receiving feedback from users is a very important role. Therefore, balancing between the ability 185‘Soolal Media and Social Networking to build automatic based web data extraction systems and human interference i one of the main targets in the field for achieving the most oat ele Another challenge is the ability of web data extraction methods to analyze large amount of data. This challenge is considered one of the most important targets especially in, the business fields (Balke, 2012). A third challenge is to maintain the user privacy in the network. Moreover, a need for having a large set of training data is one of the main challenges that should be considered, as training data needs tobe labeled, this task is usually time consuming, Finally, as web structure usually changes over time, coping with this change and finding a method for extracting information from different structure is not an easy task to be performed. TEXT ANALYTICS DIRECTIONS IN SOCIAL MEDIA Social media had its own multi-data sources which include images, video, audio, spatial data, and text, This variety gives us insights into social networks and groups that were not previously possible in both scale and extent. Unfortunately, textual data in social media presents many new challenges due toits distinct characteristics. Text analytics have challenging features of text in social media, including time sensitivity, short length, unstructured phrases, and abundant information (Hu & Liu, 2012) Text Analytics of Social Media had been growing up rapidly than traditional methods to process textual data in social media. Recently, a number of methods have been proposed to handle the textual data with new features. In this section, we introduce a variety of applying text analytics to social media as. © Event Detection. * Collaborative Question Answering © Social Tagging, Opinion Mining. © Fraud Detection. Event Detection Monitoring a data source and detecting the event that was taken within that source is, the target of Event Detection. Monitoring and tracking of news articles, digital books tesa a much attention for extracting useful information in realtime but variety Of texts in traditional media, microblogging text: i social elations Gu, ea, 2013). Ne MOWYe short and embedded with 198‘social Media and Social Networking There are four tasks related to analyzing events from social media data. The four asks are, new event detection, event tracking, event summarization, and event association. The four tasks enable the organization of methods and systems based on a task-specific view. For example, organizations could help in evaluating the methods for addressing each task and potentially, identify the missing pieces to practically tackle individual tasks when analyzing social media data (Weikum & Theobal, 2010) Considering the real-time nature of Twitter (Sakaki, et al., 2013) used Twitter users as a sensor to detect an event based on sensory observations. His aim was to ‘monitor the event such as earthquake by semantic analyses where applied to tweets. It classifies the event into a positive and a negative class...Location estimation methods such as Kalman filtering and particle filtering are used (o calculate the locations of events. (Sakaki et, al., 2013) presented an earthquake reporting system, which is a novel approach to notify people of an earthquake event. COLLABORATIVE QUESTION ANSWERING Answered to answer questions posted by other people is common define as Collaborative Question Answering suchas Yahoo! Collaborative question answering services started to develop with blooming of OSN. A large volume of questions are asked and answered every day on social Question and Answering (QA) web sites such as Yahoo! Answers, Collaborative question answering portals are a popular destination for users looking for advice with a particular situation, for gathering opinions, for sharing technical knowledge, for entertainment, for community interaction, and for satisfying one’s curiosity about a countless number of things (Liu & Eugene, 2011) (Dou, etal., 2012) proposed a graph based approach to perform question retrieval by segmenting multi-sentence questions. The authors’ first attempt to detect question sentences using a classifier built from both lexical and syntactic features, and Use similarity and reference chain based methods to measure the closeness score between the question and context sentences, On the other hand, systems provide corresponding quality QA pairs from answer's point of view. Adamic and collegues (2008) evaluated the quality of answers for specific question by analyzing Yahoo! Answer's knowledge sharing activity. First, forum categories are clustered according tothe content characteristics and patterns of interaction among users. The interactions indifferent categories reveal different characteristics. Some categories are more like expertise sharing forums, while others incorporate discussion, everyday advice, and support. Similarly, some users focus narrowly on specific topics, while others Participate across categories. Second, the authors utilize this feature to map related 157‘Social Media and Social Networking categories and characterize the entropy of the users’ interests. Both user attributes ‘and answer characteristics are combined to predict, within a given category, whether 4 particular answer will be chosen as the best answer by the asker (Adamic et al, 2012) : ee improve QA archives management, there are a number of works done by evaluating the quality of QA pairs. Harper etal. ried to determine which questions and answers have archival value by analyzing the differences between conversational questions and informational questions (Adamic et, al. 2008; Harper et, al., 2009) Social Tagging Social tagging is a method for Internet users to organize, store, manage and search for tags / bookmarks (also as known as social bookmarking) of resources online. ‘Where User-generated keywords ~ tags — have been suggested way of enhancing descriptions of online information resources, and improving their access through broader indexing. “Social Tagging” refers to the practice of publicly labeling or classification resources in a shared, online environment. Unlike file sharing, the Fesources themselves aren't shared, merely the tags that describes them orbookmarks that reference them, The rise of social tagging services presents a potential great deal of data for mining useful information on the web. The users of tagging services have created @ large volume of tagging data which has attracted recent attention from the research community (Trant, 2009). Opinion Mining “What other people think” (Lee, et al, ., 2010), one of the main critical questions in business fie Id. A question that has an answer which may change the whole plan of business and affects tremendously the decision makers’ direction. One of the main Sources to extract people's opinion is social media, therefore, many work have been performed following aspect (Agrawal, 2003), A research that focused on the characteristics of social media that can support opinion extraction, it provided 4 classification method which aim is to classify the person's opinion into negative oF positive, however, the work performed can be further enhanced by considering Active submitted opinions for more accurate results, Another work by (Sobkowiez, al. 2012)is a framework for analyzing the contents of social media with providing a ‘mining method for extracting political opinions which needs to be further extended to consider the policy-making in social media s ly ofthe impact. ae ‘onopini it nt fields including business, polities, and others, The study discusses ye ees idy discusses the social media elation 158‘Social Media and Social Networking with opinion mining and provided different approaches for extracting opinions from different social media platforms. Fraud Detection ‘There are many approaches to detecting fake or false user accounts. Some approaches attempt to systematically rank potentially fake accounts by their probability of being fake; others use supervised learning algorithms, while others still use social honey pots. One approach (Benevenuto et al., 2010) to detecting cloned profiles on different OSN involves extracting information from a real account and querying the Internet with information from that account. According results, user information is categorized as common or user specific. Account data were queried to find possible profiles that are real account. According a list of possible clones, which given a similarity score in relation to the real account. f Toward the process end, the user is ranked by similarity score, by human intervention. Also SybilRank (Cao, et al., 2012),was used to detect fake accounts by properties of the social graph to rank users by their probilty of being fake SybilRank use random walk from real account with higher normalized probility than fake account. also other approaches use social honeypots (Egele, et al., 2013). Honeypots are created to trap attackers and begin monitoring and logging attackers’ activity. When a honeypot’s profile receives an unexpected friend request, the user sending the request is put under observation. The user's activity is tracked for later use as evidence by a classifier to decide if the suspected user is a spammer or not. In this way, the authors identify spammers with a low false positive rate and are able to identify characteristics of spammers’ profiles that can be applied to detecting previously unknown spammers. Elazab and collegues (2016).present fake accounts detection in Twitter based on minimum weighted Feature. The minimum set of attributes for detecting the fake accounts on Twitter has been determined and tested. Five of the best classification algorithms have been applied and the results have been compared 3. Evaluating both steps is applied and compared with other researchers’ results which proved the advancement in the accuracy level of the proposed approach (Elazab, et al, 2016) The final class of approaches to detecting spammers assigns users a vector of Values (“feature” values) (Camasani &Calzolari, 2012; Martinez-Romo & Araujo, 2013; Yang et, al. 2010; Stringhini, et al. 2010) capturing different attributes of their Profiles, Tweet history, local social graph, etc. Features are carefully constructed Using empirical user data, before they are used as input to a supervised machine learning algorithm along with a set of users that have been pre-classified as being Spammers or not. The resulting classifier can then be used on the broader social network. Approaches based on machine learning can be augmented using statistical 159‘Social Media and Social Networking sed in Tweets (Kontaxis et, a, 2011). For example, a ding topic can be compared to the broader thread of of Kullback-Leiblerdivergence. Augmenting typical profile-based features with technique ‘Tweet-based. features can improve the performance of a spammer-cassfication system. Detecting subversion on twitter cour work uses a machine learning framework with profile and Tweet based features, We synthesize diverse ideas from the literature and describe our proof of-concept system that samples the Twitter social, computes interesting features on the data, and classifies users as being fake or not. So far, social networking websites lack automated systems to detect fake accounts because it is very difficult to monitor of fake and real online social network Profiles Using hacked accounts to send spam can be viewed as an instance of a Sybil attack} on a social network. “There are two types of Sybil attacks: one involves creating many fake accounts, and the other employs hijacked legitimate accounts to perform a desired function, the first attack is common for Twitter,. An attacker may respond to such a defense b hacking into legitimate users’ accounts and using them to carry out their attack. The ‘compromised accounts, being legitimate forthe most part, are notreadily Identifiable, as their behavior is less divergent from that of a real user (Benevento, et al., 2010), analysis of the language us suspicious Tweet about a tren ‘Tweets about the topicusing the concept: Issues of Textual Data Analytics in Social Media ‘The analysis of the new features of textual data in social media from four differen perspectives: Time Sensitivity, Short Length, Unstructured Phrases, and Abundant Information (Hu & Liu, 2012). Time Sensitivity ‘Time Sensitivity is important clement of numerous OSN administrations is the real-time nature, bloggers examine their sites like clockwork, Which they po news and data a few times day by day. Clients may need to speak in a flash wit ‘companions about “What are you doing?” (Twitter) or “What is at the forefront your thoughts” (Facebook). At the point when presenting a question to Twitter, returned results are just a few minutes old (Sakaki et al., 2013) Short Length Short Lengthis one of the imitations in OSN sites.as they more often than not conti the length of clients’ made substance, for example, micro blogging messages, ite 160 a‘social Media and Social Networking audits, QA sections and image inscriptions, and so forth. Twitter permits clients to constrained to 140 characters. Likewise, Picasa remarks are constrained to 512 characters, and Windows Live Messengerare confined to 128 characters. Information witha short lengthis universal on the web at present. As aresult, these short messages have assumed expanding vital parts in applications of OSN. Fruitful preparing short messages is essential to content examination routines. (Cai et al., 2008). Unstructured Phrases the quality of the content is different between text in osn and traditional media, The variance of quality originates from people's attitudes when posting a microblogging message or answering a question in a forum. Some users are experts for the topic and post information very carefully, while others do not post as high of quality. The main challenge posed by content in social media sites is the fact that the distribution of quality has high variance: from very high-quality items to low-quality, sometimes abusive content, this makes the tasks of filtering and ranking in such systems more complex than in other domains (Ventola, 2014) Moreover, when composing a message, users may use or coin new abbreviations or acronyms that seldom appear in conventional text documents. For example, messages like “How r u?”, “Good 9” are not really words, but they are intuitive and popular in social media. They provide users convenience in communicating with cach other, however itis very difficult to accurately identify the semantic meaning of these messages. Exploiting the Power of Abundant Information Online networking when all is said in done show a rich assortment of data sources. In expansion to the substance itself, there is a wide exhibit of non-content information accessible. For instance, Twitter permits clients touse the”#” image, called hashtag, to Stamp decisive words or points in a Tweet (tag data); a picture is normally connected With numerous labels which are described by diverse districts in the picture; clients, ie able to fabricate association with others (interface data) in Facebook and other interpersonal organization locales; Wikipedia gives an effective approach to users to Sidetrack to the equivocalness idea page or larger amount idea page( semantic pecking order information). All these outer data presents open doors for traditional tasks. Past content examination sources dependably show up asstructure, hile the content investigation in online networking has the capacity infer data from ‘ifforent perspectives, which incorporate client, substance, connection, label, time Samp and others (Battaglino et al., 2013) 161KEY TERMS AND DEFINITIONS Facebook: Is an American for-profit Corporation and online social media and social networking service. LinkedIn: Is a business and employment-oriented social networking service that operates via websites. Orkut: Is a social networking website owned and operated by Google to help users meet new and old friends and maintain existing relationships. OSN: Orbit Showtime Network (stylized as “osn”)-- Is adirect-broadcast satellite provider serving the Middle East and North Africa (MENA). It offers popular entertainment content such as movies, sporting events and various TV shows from major networks and studios. Social Media: (Also social networking site or social media) is an online platform that is used by people to build social networks or social relations with other people who share similar personal or career interests, activities, backgrounds or real-life connections, and where can purchase online. ‘Twitter: is an online news and social networking service where users post and interact with messages, “tweets,” restricted to 140 characters. Registered users can Post tweets, but those who are unregistered can only read them. Users access Twitter through its website interface, SMS or a mobile device app. Web 2.0: describes World Wide Web websites that emphasize user-generated content, usability (ease of use, even by non-experts), and interoperability (this means that a website can work well with other products, systems and devices) for end users. Wikipedia: Is a free online encyclopedia that aims to allow anyone to edit articles. It is the largest and most popular general reference work on the Internet and is ranked among the ten most popular websites. 167