Social Media in Business and Society

Most organizations have a tendency to look upon interpersonal advertising as a threat, where some even choose to ban the use from the work area altogether. The idea behind it being that employees would be given the possibility to spend your time online, chat, and possibly create as a security risk to the organization. (Turban, 2011)

(Smith, 2010) outlines threat of employees' social marketing use at the job, these can be both intentional or not and they could lead to legal and reputational risks for organisations. These have been categorised as three main problems
  1. Use of communal media cannot be fully regulated, checked or manipulated thus organisations are "quitting control".
  2. Social marketing is a worldwide means of communication, once a negative post is online it's only a matter of their time till it moves viral thus attaining competition, regulators and customers.
  3. Social marketing is mental and employees can point out their feelings of contentment and/or annoyance.

Furthermore, (Flynn, 2012) recognizes the risks of having employees taking part in social mass media by "causing reputational damage, result in lawsuits, cause humiliation, crush trustworthiness, destroy jobs, create electric business data, and lead to productivity loss".

(Dreher, 2014) argues that interpersonal media is never to be feared, but instead embraced and viewed as a chance where employees can become "corporate advocates and brand ambassadors". If anything, it helps employees keep up thus far with latest news related to the industry together with continuous knowledge development. Nonetheless, even though there are many studies that point out the advantages of social multimedia, there continues to be no clear-cut decision whether it can effect work performance or whether it can gas the interpersonal capital of the employees and assist in knowledge copy (Zhang, 2016).

However, it can't be denied that each organisation allowing interpersonal media at work will always have its fair package of troubles to conquer. (Eliane Bucher, 2013) speaks about medical issues that can be came across. Starting with stating that there is very much information on social media that pros may face information overload. Not to mention the mixture of work life with private life overlapping with interpersonal media.

New technology should improve worker's efficiency and reduce stress levels however usually the other occurs (Eliane Bucher, 2013). Technostress as described by (Brod, 1984). To be successful in the interpersonal marketing environment one needs to overcome the below 3 factors usually technostress is produced
  • Techno-overload - Upsurge in workload that could be actual or recognized.
  • Techno-invasion - Social media enables visitors to be constantly linked from nearly every device. This may lead to the sensation of the "need" to be linked or online triggering reduction in family time allowing work issues to invade the private life (Eliane Bucher, 2013).
  • Techno-uncertainty - Social media is constantly changing and therefore brings with it uncertainty as regards to what systems and skills are needed to perform the job and exactly what will they maintain the future.
Social media comes with many legalities linked with it. These range from pre-employment to create employment. Wrong consumption of social mass media will for sure lead to waste products of time, inefficiency, reputation issues and negative image for the organisation. A number of the laws are layed out below by (Lieber, 2011)

Employment Laws by tagging co-workers in certain provocative photographs or videos,

Defamation and Libel Laws by proclaiming certain remarks on co-workers or employers thus effecting their reputation. , As stated in (Trott, 2009) a Microsoft Study found that 41% of employers structured their decision of not hiring an applicant based on what they found online with regards to their reputation. This is also known as "Netrep". This takes its legal threat of discrimination alone if the recruiter is basing decision on the netrep.

Fair CREDIT SCORING Act with interviewers "friending" a job candidate on Facebook to acquire more information than is required for the job applied.

Health Insurance Portability and Accountability Act by having a medical professional LinkIn with a patient.

Uniform Trade Secrets Act by having employees talking about or commenting on communal multimedia about company "internal only" conversations or non-public tasks.

Employers can keep an eye on the use of social marketing at work if the employees are prepared beforehand. Disciplinary activities can be taken once any misuse is being noticed. Policies will include what is allowed and what is considered as abuse (Trott, 2009).

If the employees post on the personal accounts beyond office time and such posts are with regards to work having a poor impact for some reason to the company or the company then there continues to be grounds for disciplinary action even though employees try to advocate for "respect for private and family life, home and correspondence" (article 8) or "freedom of expression" (article 10) from the Individual Rights Act 1998.

As discussed above, social media has its benefits and drawbacks and seeing that social media is here to stay organisations have little options but to simply accept the new actuality, treat it and understand how to make good use of it. (Lieber, 2011), amongst others, identifies the next standards that any company willing to harness social multimedia must treat
  • The creation and enforcement of sound social media policies within the organisation's employees addressing good use, access during work time and basic behaviour on public multimedia (even during personal time).
  • Directly using sociable marketing for the benefit for the organisation such as for recruitment, marketing and looking into competing organizations.
  • Monitoring of key internet sites to data mine information relating to your organisation (and probably others' as well), possibly using automated algorithms and software for maximum efficiency and exactness.

From the above-mentioned requirements, the first two package with "human tool" aspect of social media where organizations construct guidelines with their employees about how to use them, and they as the organizations may use social media directly for recruitment, marketing etc. However, as the third criteria advises, to make most use of internet sites organizations must ensure that any information/data being released on such websites, is compiled and used effectively.

It is important an organization is always alert to what the average user says about their brand, effectively getting the overall feel or mood while analysing the trends across time. The exact same principle could be employed to monitor competitors; possibly for example identifying any vulnerable products which the competition have and having your own similar product take benefit of the situation.

Effective monitoring originates from creating good data. Data mining involves the following steps to make data important for monitoring: (Raghav Bali, 2016)

  1. Removing unwanted data and noise
  2. Transformation of the natural data into data you can use for further processing
  3. Study the info and produce patterns that can give further insight to your data
  4. Represent the data in a manner that pays to to companies or even to who the info designed for.

There will vary data mining techniques which can be used to screen social media use. Social media is a form of real time communication therefore an efficient monitoring tool needs to monitor and offer alerts as things happen. Most content material mining tools employ search engines to go through social marketing sites and acquire information related to the keywords or passions. (Make My Words article)

Text Analytics (Text/Data Mining)

Text analytics involves a complex and elaborate amount of steps to strip down conversations into independent words and analyse the way these words are being used, positive or negative and even derive habits from accumulated data.

When we search for a movie and acquire various other movie suggestions that technique is using word mining.

Text Mining is made up of Data Mining (Information retrieval, Natural Dialect Processing & Machine learning) + Words Data (Emails, Tweets, Media Articles, Websites, Sites etc. )

Figure 1: Content material Mining (Charu C. Aggarwal, 2012)

As indicated in Body 1, Stop Term Removal and Stemming eliminate the common and less important words form a saying, this helps categorizing different words with same meaning as "see", "seen" and "being seen".

Bag of Words (BOW) is having words segregated from the phrase and each phrase getting a numerical value which presents its importance.

Limitations

(Charu C. Aggarwal, 2012) outlines several limitations that can be noticed and future in-depth research is required
  1. The real-time posts on social media are a very important source of information as mining data in real time as it has been posted can deliver many advantages. This however remains a challenge for when these posts are not conducted from work personal computers or from outside the house work.
  2. Social advertising is very unstructured and some applications like tweets even limit the quantity of characters per post. This brings about problems of wording recognition when brief length words are used like "gnite" "gr8" etc.
  3. Social media allows various ways to express thoughts or thoughts these could be through images, videos and tags making the written text analytics a lot more complicated and difficult in its pre-processing stage.

Method 1:

Keyword Search (Rappaport, 2010)

Organisations can make a decision which keywords they want to monitor, these may be chosen predicated on what is important for that company, it could be their products or mental states. Social networking is an extremely unstructured place including sound and unwanted data for our data mining process. This form of search is good to fully capture keywords and try and form a interpretation of the words and the occurrence used however it's very hard to create what's the user's objective. Because of this, we then look at a more technical search method called Sentiment Evaluation.

Method 2:

Sentiment Evaluation and Emotion Analysis

Sentiment Analysis is the procedure of identifying sentiment in text message and analyse it. You can find three types of sentiment evaluation (Walaa Medhat, 2014)
  1. Document Level - Analyse the whole document as one matter and form an judgment or sentiment on the complete document
  2. Sentence Level - Analyse sentiment in each sentence
  3. Aspect Level - Analyse sentiment according to entities as you could have several aspect in a phrase for the same entity.

For this research, we are centering our research on Word Level research using Semantic search.

Semantic Search:

Semantic search should go beyond the traditional keyword search by giving a interpretation to a term and employs an array of resources to interpret the word and thus providing a far more accurate consequence.

Some types of semantic search in our daily lives
  • Conversational searches:

Figure 2: Conversational Search (Yahoo, 2017)

  • Auto Right spelling faults:

Figure 3: Vehicle Correct (Google, 2017)

  • Display information in design format:

Figure 4: Information in design format (Yahoo, 2017)

(Charu C. Aggarwal, 2012) outlines some troubles that are encountered when going through mining. These are the difficulty in recognising opinions, subjective phrases and thoughts.

Opinion mining obstacles.

When using semantic search method on the post one must understand that the post can contain all the next:

  1. Positive viewpoints "I love the computer I purchased, it has an extremely clear screen"
  2. Negative thoughts "however my partner thinks it's very costly"
  3. Different goals "The focuses on in the positive viewpoints relate with the computer and the display screen whereas the goals in the negative views will be the price"
  4. Different opinion holders "The positive opinions are mine nevertheless the negative opinions are of my partner"
Subjectivity mining challenges

Posts are also composed of target and subjective responses. Subjective expressions like thoughts, desire, assumptions amongst others may well not contain viewpoints or might not exactly express any positive or negative commentary.

Emotions mining challenges

Emotions (love, happiness, anger, dread, sadness, happiness plus more) fall under a kind of subjective appearance. Sometimes thoughts give no thoughts in a phrase.

To take notice of the effectiveness and ideal procedure towards the research of social marketing related articles and messaging, a software algorithm was designed and partially developed to demonstrate this scenario. The theory behind this software is to really have the user write inside a textbox, mimicking an actual employee typing by using a company machine, as the system screens such words and works per what it registers. Therefore, this tool will be offered as a standalone software/algorithm notion, emulating an actual activity of a possible employee, and as such must be designed accordingly to make use of it in a real-life situation.

The basic theory of the solution proposed is made up of three modules
  1. The key logger that displays the user's type at runtime and results certain rules
  2. The keyword and semantic research on the info gathered
  3. The storage of produced examination and log

The following flowchart outlines the lifecycle of said solution, followed by a detailed evaluation of each part mentioned above, as well as it can be ways how it could be further enhanced to create even more correct results.

The movement of the proposed solution.

Created using pull. io (https://www. draw. io/)

Collecting and Processing Data

In this solution, key logging is utilized to monitor the data inputted by the user, which is a constant monitoring of the keystrokes authorized by one's activity, and listed as a blast of text ready to be dissected and analysed as required. The primary advantage of using such a technique is that data is accumulated and used in real-time, making it ideal for cases where an security alarm (for example a negative post related to work) needs to be raised as fast as possible to the relevant workers, providing a detailed log of the actual staff has typed (through the key logger) eliminating the need to monitor and gain access to the relevant public media to check what has been posted.

Note: there are other strategies you can pursuit to screen the user's activity, such as firewall insurance policies or standard network surveillance, however in real-life situations such solutions can prove rather difficult to create because of the expertise required; while web encryption and proxy services helps it be even harder to effectively keep an eye on the traffic generated by the users.

A key logger, even if effective, produces a lot of unneeded garbage beyond the opportunity of social press. For example, a worker focusing on his station would be constantly registering keystrokes that your logger is then adding them up to its text stream. This could end up being very difficult for three main reasons
  1. The logger would start to amass a significant amount of space for storage, unless the main element logger is given a limit of how much information it can hold and removing "old" data to make up space for the "new" data, but than some information can get completely lost.
  2. The research of the text stream generated could be very intense, which can significantly impact the performance of the device doing the examination, especially when considering that the research is assumed to be processed on the user's machine which almost certainly isn't perfectly suited for such intensive work. Furthermore, following a prior point, the "garbage" log is being analysed too needlessly.
  3. The chances are that an staff would spend hardly any time on communal multimedia, thus logging and analysing the work-related activity is quite pointless for such a range.

To triumph over the above-mentioned issues, the proposed solution employs predefined social marketing result in keywords i. e. a set of social multimedia websites such as Facebook, Twitter, LinkedIn etc. , where depending on such triggers being strike or not, the key logger will have two states, passive monitoring and energetic monitoring.

When the tool is jogging normally, the key logger is a passive state keeping only the last 30 heroes in its storage, without finalizing the stream. The only thing it can however, is to constantly check the stream read from the textbox in the tool against the trigger keywords, in case any of the keywords is found to get been registered then your key logger would go into active state. While in this point out the key logger would increase its maximum capacity, and commence to log every keystroke while constantly analysing the give food to. The key logger will get back to passive point out when the predefined personality limit is reached or plenty of time has handed.

Following this reasoning, only a couple of keystrokes would be signed up, reducing the opportunity of collecting and processing unneeded information while keeping the workload and storage space use of the device to the very least.

Note: in this process after the key logger switches into active condition, it is monitoring and analysing the supply at runtime locally, which could prove to be quite intensive with respect to the parameters established and the entire performance of the users' machines. Organizations utilizing this solution can choose to hold the log analysed following the key loggers dates back into passive status and therefore analysing the data only once. Better yet, since the solution assumes that the main element logger is analysing the data locally, instead the logs can be delivered to the server and be analysed as a scheduled task.

Once this data is captured through the key logger the give food to can be processed by means of the methods mentioned earlier (Method 1 and 2). Predicated on the results we store the info in our information system and align the data based on the organisation's interpersonal policy.

Approaching data evaluation using keyword and semantic methods

The designed software makes use of two different types of analysis algorithms, keyword based mostly and semantic founded, and are being used together to cancel each other's constraints and thus providing much more exact results.

Keyword structured analysis

The more traditional keyword research algorithm involves having a list of keywords i. e. a predefined group of texts, and hit the data to be analysed against that list to ascertain whether any keywords have been strike and at what frequency. For instance, having a text (representing the info) analysed against a list of negative texts (the keywords) would provide a group of statistical information which could be used to judge how negative the written text is, which is conceptually what a social advertising monitoring tool should be striving to attain.

However, the major flaw of this examination algorithm within the context of social marketing monitoring, is the fact that keyword based research is much too broad and susceptible to wrong alarms if not operated. Having the data gathered from the key logger (therefore filtered to communal marketing activity) analysed against a couple of negative texts, the statistical information produced might not be relevant to the organizations' interest. A worker could simply be placing a feed about how bad the elements is and exactly how much s/he hates it, that your keyword examination algorithm would discover as negative and article accordingly.

In the proposed solution, the keyword established algorithm uses two different packages of keywords contrary to the accumulated data, with desire to to filter the batches of logged text messages by relevance. The first place includes a set of works related wording, such as 'work', 'job', 'company', '[company name]' etc. i. e. every keyword which could somehow link the user to the business implementing the perfect solution is. In the next set, a set of keywords/texts associated with negativity are stored, such as 'bored', 'disappointed', 'hate', 'flat', 'suffering and fatigued' etc.

When the info approved along through the key logger reaches the keyword evaluation module, it would first check the log from the first set and for that reason determine if the data fed is of any relevance to work, and if not only do nothing. Alternatively, if the keywords from the first collection is hit, this means that the info inputted is pertinent and for that reason must be analysed further. In cases like this, the tool would analyse the whole log within the main element logger (which happens to be in an energetic state as referred to in the previous section) and draw out the statistical information based on the second set in place.

The circulation of the entire keyword structured algorithm designed in the tool

Created using sketch. io (https://www. draw. io/)

Examples

Keywords to suppose

First Collection (work): 'WORK', 'JOB'

Second Set (negative): 'Bored stiff', 'UNHAPPY', 'SAD', 'HATE', 'DULL', 'TIRED', 'SICK AND Fatigued', 'ANNOYED', 'FED UP'

Example 1

Input

Hate this weather, it's severely effecting my ambiance. Constantly feeling tired and miserable.

Output

None

Example 2

Input

At work and tired. Wish I could find a much better job, this one is just so irritating.

Output

BORED x 1

[full log]

Example 3

Input

Never a dreary moment at the job. By the end of your day, the management brought in pizzas, fresh doughnuts and ale. In a couple of hours, the meals was gone going out of everyone too worn out to move. Got to love the corporation, always making sure their employees are never bored and unhappy.

Output

DULL x 1

TIRED x 1

BORED x 1

UNHAPPY x 1

[full log]

From the samples above one can be aware a few restrictions concerning the keyword based analysis algorithm.

In example 2 the logged words is alarming, which almost certainly would require the entire attention of the responsible personnel, but because of the limited keywords, only a single piece of text was hit which would make the productivity seem not so alarming. Furthermore, the logged content material had the word 'annoying' which in the negative keyword place is posted as 'annoyed', but nonetheless this is not captured. Therefore, this means that this algorithm is highly dependent on the keywords lists and possible deviations of each text.

In example 3 the productivity appears very alarming since the negative keywords list was hit 4 times, but the type is very positive. The algorithm was unable to take into consideration the framework of the way the negative words were used and counted the amount of times they were came across within the log, hence nurturing a false alarm.

To overcome such restrictions, other algorithms must be used in conjunction with the keyword established, where in this solution the semantic established approach can be used to go with the algorithm and make an effort to provide more accurate results.

Semantic based mostly analysis

As described in previous parts, semantic analysis introduces a certain degree of understanding when analysing a given text, and this is attained by giving meaning from what it is fed. In this suggested software algorithm, this kind of analysis is utilized to judge the sentiment and feelings behind the fed input, and therefore can determine whether the users' work related activity on cultural media is negative or positive, which by expansion might be able to overcome the constraints of keyword centered approach.

Basic types of semantic based mostly algorithms used to analyse words in relation to sentiment and feeling, often providing an individual value output denoted by a share, where 0% means that the written text is absolutely negative and a 100% would suggest that without a doubt it is positive. However, semantic evaluation is capable to go beyond a simple value, where a few of which can create a fully detailed record indicating the amount of feelings for multiple types, such as anger, fearfulness and delight. The next is an exemplory case of such a written report produced by the tool Shade Analyser provided by (Cloud, 2017).

Example report of your semantic established algorithm made available from IBM Watson Programmer Cloud

Applying this algorithm which produces an extremely detailed survey, may be well beyond the opportunity of monitoring work related activity on social media. In the end, what the suggested solution is wanting to achieve is to detect negative activity which would harm said organizations', that when recognized, the log of that activity is passed along to the equivalent staff with perhaps a brief article of the examination.

Another downside to be considered in this scenario, is that light weight semantic algorithms are much less intense than algorithms which consider different kinds of feelings when analysing a content material, and considering that in the perfect solution is such an analysis will be prompted almost constantly, having much algorithm being triggered would lead to an extremely negative experience to said users. This is why in the suggested solution a lighter semantic examination is considered, that is the API provided by (ParallelDots, 2017).

Note: one could argue that using a semantic evaluation algorithm which produces an in depth record, could replace the complete algorithm which is using both keyword based evaluation and the light weight semantic based evaluation. However, performance wise the last mentioned would operate much better, and from a specialized perspective considerably easier to setup.

Note: in the proposed solution, the semantic examination will be conditional to if the keyword structured algorithm is triggered or not, and for that reason at the mercy of the filter which is discovering if the activity on social media relates to work or not.

Examples using the sentiment analysis demo provided by (ParallelDots, 2017), which outputs single value percentages 0% being negative, while 100% being positive.

Example 1

Input

Hate this weather, it's greatly effecting my ambiance. Constantly feeling worn out and unfortunate.

Output

0%

Example 2

Input

At work and bored stiff. Wish I could find an improved job, this one is just so annoying.

Output

6%

Example 3

Input

Never a uninteresting moment at the job. By the end of the day, the management earned pizzas, fresh doughnuts and beer. In a couple of hours, the food was gone going out of everyone too worn out to move. Got to love the corporation, always ensuring their employees are never bored and unsatisfied.

Output

79%

Classifying severity predicated on score and frequency of words

Thus very good, the algorithm recognized negative activity on interpersonal media associated with work, using both keywords and semantic evaluation. However, the term negative can be alternatively broad and it can be the situation that the organization would not want to be alerted for each small negative activity, since that will become counterproductive. Consequently the proposed algorithm has a threshold mechanism which establishes whether to submit alerts or not.

The threshold settings are two. The minimum number of negative words the activity must contain, and the bare minimum percentage of negativity to be considered. Right after the main element logger is finished monitoring the communal media activity, if work related activity is logged, the system evaluates the log predicated on the threshold arranged by the administrators of the machine, and proceed appropriately.

Using same guidelines of previous example for keyword and semantic founded techniques. The thresholds are arranged as follows: Minimum amount Keywords 1, Least Semantic Percentage 30%.

Example 1

Input

Hate this weather, it's significantly effecting my feelings. Constantly feeling worn out and miserable.

Output

None (not work related)

Alert

No

Example 2

Input

At work and tired. Wish I could find a much better job, this one is merely so troublesome.

Output

Keywords struck: 1

Semantic: 6%

Alert

Yes

Example 3

Input

Never a lifeless moment at the job. At the end of the day, the management earned pizzas, fresh doughnuts and beer. In a couple of hours, the food was gone departing everyone too worn out to move. Surely got to love this company, always making certain their employees should never be bored and unsatisfied.

Output

Keywords struck: 4

Semantic: 79%

Alert

No

Limitations of the existing approach

The software that has been designed and partly developed as a concept for this paper must be further increased and developed to be able to be turned into a fully communal marketing monitoring system in the task place environment. The suggested solution because of this paper has several limits that for the range of this newspaper can be bypassed however for an effective implementation they would have to be addressed
  1. The first limitation is that the machine produces data which currently needs to be analysed by the person, even if different filters and threshold system are applied at tips through the process. The individual in charge must monitor the text being captured and make a decision if this is of relevance to public media publishing and if it's of any harm to the company or its employees. This will be beat with having a Decision Support System integrating with the info being collected by our monitoring software.
  2. Another limitation is that currently it will start recording data once it detects certain keywords in the written text. This will be in conjunction with the company firewall that once certain urls are diagnosed the monitoring plugin kicks in and can stop monitoring once the social mass media sites are not active.
  3. When the program detects anything it presently does not stop an individual from posting nonetheless it logs the info in the information system database for this to be analysed. This should be enhanced in a manner that an individual is alerted before posting the current post if it appears to violate the organizational sociable policy and preventing it from going viral.

Effective execution & Future Enhancements

The software should be converted to have the ability to be installed as a plugin for browsers. Every time the browser loads up the plugin would stimulate and starts monitoring the urls that the user is launching within the web browser. After the plugin picks up certain predefined urls of public mass media like Facebook, twitter etc. the plugin begins collecting data.

Companies would have to have certain procedures in place

  1. A policy as to which are the approved browsers to be utilized within the company and effective constraints should maintain destination to avoid users setting up other browsers. This would enable the plugin to be installed on typically the most popular browsers.
  2. A insurance policy on permissions which should not allow users to modify the internet browser plugins by disabling or uninstalling the monitoring plugin.
  3. A insurance policy that informs users that personal computers at work are subject to monitoring which any misuse of such equipment or using the mentioned equipment to damage the reputation of the business or its fellow workers would engage in disciplinary actions.

Once data is collected the system should store this data in an Information System so the accumulated data can be analysed and acted upon. Records should be generated when certain thresholds are met.

Even though we'd be monitoring any communal mass media used from work computer systems there is a restriction to the proposed solution that is we will never be catering for Mobile devices, even if they're on our work network infrastructure. Even more mobile devices connected using a 4G data plan are not being checked. For effective monitoring of mobile devices on the task infrastructure an alternative approach must be looked at.

Other approaches that may appeal to some limitations of this approach are real-time search engines like google styles, and tools according to (Home, n. d. ). These tools can be used when the organization needs some degree of off premises monitoring, they can provide indications as to what has been said by everyone on social mass media which is often used as a sign or added information from what is being gathered on premises.

References

Brod, C. , 1984. Technostress: the individuals cost of the computer revolution. In: s. l. :Addison-Wesley.

Charu C. Aggarwal, C. Z. , 2012. Mining Words Data. s. l. :Springer Research & Business Press.

Cloud, I. W. D. , 2017. IBM Watson Designer Cloud. [Online]

Available at: https://tone-analyzer-demo. mybluemix. net

[Accessed 3 3 2017].

Dreher, S. , 2014. Social websites and the world of work. Corporate and business Communications: An International Journal, 19(4), pp. 344-356.

Eliane Bucher, C. F. &. A. S. , 2013. The Stress potential of social media at work. Information, Communication & Modern culture, Volume level 16:10, pp. 1639-1667.

Flynn, N. , 2012. The SocialMedia Handbook: Insurance policies and Best Practices to Effectively Manage Your Organization's SOCIAL NETWORKING Presence, Articles, and Potential Risks. Pfeiffer, San Francisco, CA.

Google, 2017. Yahoo Search Engine. [Online]

Available at: www. google. com

[Seen 3 3 2017].

Home, H. , n. d. Howards Home. [Online]

Available at: https://www. howardshome. com/

[Accessed 19 2 2017].

Leftheriotis, I. &. N. G. M. , 2013. Using social mass media for work: Getting rid of your time or improving your work?. Division of Informatics, Ionian School, Section of Computer and Information Technology, Norwegian University of Research and Technology. .

Lieber, L. D. , 2011. SOCIAL WEBSITES at work - Proactive Protections for Employers.

ParallelDots, 2017. ParallelDots. [Online]

Available at: http://www. paralleldots. com/sentiment-analysis

[Accessed 3 3 2017].

Raghav Bali, D. S. , 2016. R Machine Learning By Example. s. l. :s. n.

Smith, N. W. R. a. Z. C. , 2010. SOCIAL NETWORKING Management Handbook: Everything You Need to learn to Get Social Media Working in Your Business. John Wiley & Sons Inc. , Hoboken, NJ. .

Trott, L. , 2009. Social Media at work.

Turban, E. B. N. &. L. T. -P. , 2011. Business public networking opportunities, adoption, and risk mitigation. Journal of Organizational Processing and Electronic Business, pp. 21, 202-220.

Walaa Medhat, A. H. H. K. , 2014. Sentiment analysis algorithms and applications: A survey. Ain Shams Anatomist Journal, 5(4), pp. 1093-1113.

Zhang, X. C. X. G. D. V. X. , 2016. Discovering the impact of social marketing on employee work performance. Internet Research, Level 26, pp. 529-545.

Also We Can Offer!

Other services that we offer

If you don’t see the necessary subject, paper type, or topic in our list of available services and examples, don’t worry! We have a number of other academic disciplines to suit the needs of anyone who visits this website looking for help.

How to ...

We made your life easier with putting together a big number of articles and guidelines on how to plan and write different types of assignments (Essay, Research Paper, Dissertation etc)