Web Social Media: Twitter Assignment Sample

Introduction of Web Social Media: Twitter Assignment

Get free written samples from subject experts and Assignment Writing in UK.

Trending Tweets

Twitter doesn't only count the number of times a phrase occurs in a tweet to decide which are the most popular trending topics. The good morning tweets and such words would gain a regular spot in the trending topics if that were to happen. In addition, the trending topics would soon turn into a spam slingshot since twitter bots may create a high number of the identical tweets in a short period of time. There are many additional elements at play when it comes to establishing what subjects are currently trending, apart from the sheer number of people involved. Algorithms are employed if a large amount of data must be sorted out using a combination of criteria with varying weights.

Twitter's search algorithm has a similar structure to Google's, with a plethora of phases. The weights and elements that go into Twitter's algorithm for trending topics are also kept under wraps, much like Google. If there is no official information, one might draw some conclusions from historical patterns. As an example, the name "WikiLeaks" is intriguing. No mentions of WikiLeaks or its enigmatic founder Julian Assange made it onto Twitter's Trending Topics list after the revelation of sensitive material.

"An algorithm recognises the top subjects that are presently being tweeted more than before," Twitter said in response to first reports of censorship on the platform: While WikiLeaks was often cited in tweets, it had a subliminal dynamic that did not influence the algorithm's trend detection. To qualify for inclusion in Twitter's trending topics, the usage of a phrase must grow exponentially, rather than just rising in popularity. When Twitter first launched, retweets, or the simple repeating of a tweet by other users, were essentially developed by the Twitter community and only formally embraced by the company later.

An intriguing piece of material may be tagged as "worth reading" and then shared with your own followers by retweeting the post. The retweets were recognised as a key role in the formation of hot topics for this same reason: The retweeter's followers get the word virtually exactly as it was first tweeted. Retweets of popular topics accounted for 31% of all tweets in the research. However, it is theoretically conceivable for a single Twitter user to start a trend on the social media platform. When comparing conventional media to social media, the dissemination of information was significantly different. Trend setters may be divided into two categories.

Twitter users who retweet a word in order to disseminate it may be considered as both the originator of the trend and the beginning point for the critical mass of mentions that subsequently occurred. Many firms and brands are likely to find these Twitterers, who are commonly referred to as "influencers," very fascinating. Retweet rates tend to be higher for those with large followings who have a great degree of confidence in them. To be fair, it's not the amount of followers, but the number of tweets a person has on a certain issue, that matters most. However, even among the top 10 Trending Topics, there are distinct variances in quality. This is a measure of how long people spend on the most popular subjects, and it varies widely.

The average issue does not stay in the top ten trends lists for more than 20 to 40 minutes, despite the fact that you may get the impression that certain subjects go on all day or even longer. Only a few themes are able to go on for an extended period of time. In part, the researchers at HP believe that the short lifespan of Twitter's trending topics is due to the intense rivalry amongst the many phrases being used. Some trends, on the other hand, "migrate" back into the top 10 after a few hours because they are popular across the globe. Long-term trending themes have a large number of "unique writers," which may have a role here. A single or a few social media accounts aren't going to help you get your message out there. To get as many people on Twitter to utilise a certain word is much more significant.

Sentiment Analysis

The hedonometer, a project of the University of Vermont's Computational Story Lab, is one of the best-known online instances of sentiment analysis. More than 50 million tweets in English are sent there every day (which is about a tenth of all Twitter message traffic). A "Happiness Index" is calculated each day based on this data. From a strictly mathematical standpoint, the strategy is straightforward: The team utilised Amazon's Mechanical Turk service to have individuals assign a "lucky note" to each of the 10,000 most often used phrases. Filtering of words that are neither neutral nor very context-dependent is done.

Each person's score is multiplied and averaged to arrive at a daily happiness score for the group. The project's website has a list of words, as well as assessments (in English and nine other languages). As a senior research analyst for Conversational AI at IDC, Hayley Sutherland describes it as an old-fashioned approach to sentiment analysis. According to the expert, this may still be ideal for really big volumes of text. Positive-negative sentiment analysis is the most prevalent sort of analysis used by the hedonometer.

Rather of using a one-to-nine scale like the hedonometer, some systems utilise a percentage or assume three values (positive, negative, and neutral). "It depends on the instrument," Sutherland explains. Sad, furious, and thrilled are some of the most prevalent feelings. ' Companies may use this detailed type of sentiment analysis to determine whether or not individuals are annoyed or irritated by the situation. It's about figuring out what the consumer wants to do," a sentiment analyst says. Sutherland cites the example of sales: "Are they interested in purchasing or not?" When it comes to sentiment analysis, you can do much more than merely evaluate text. Facial and speech analysis are used by certain systems, while others employ both.

Emotional AI is becoming more and more popular. Sarcasm, for example, may be more easily detected if the speaker's tone of voice is taken into consideration. Either statistics or machine learning may be used as a starting point for sentiment analysis (based on supervised or semi-monitored learning algorithms). Supervised learning, like the hedonometer, relies on humans to assess a collection of data.

Using semi-supervised learning, automatic learning and frequent evaluations guarantee that the system produces accurate results. Using deep learning, "Deep learning employs multi-layered neural networks that are based on the functioning of the human brain," says Sutherland. This advanced level of sentiment analysis can scan full words or complete conversations, as well as voice recordings and video material, to detect emotions. All major cloud providers as well as companies providing customer support and marketing solutions now have sentiment analysis capabilities available.

Boris Evelson, Principal Analyst at Forrester Research, advises companies interested in sentiment analysis to first look at the tools and technology they presently employ. Does their survey software offer sentiment analysis??" IBM Watson Discovery and Micro Focus IDOL, as well as customer feedback management platforms, provide sentiment analysis possibilities. We suggest our clients to check around first, since they frequently require sentiment analysis as part of the document gathering and review or the customer experience process." " Only a few firms have built their own sentiment analysis software. Both internal knowledge and extensive training data sets are needed for this.

If your company has extremely particular needs that aren't being covered by current platforms, it may be worth your while to make the investment. If this is the case, open source libraries are commonly used as the foundation for self-created tools. According to Dan Simion, vice president of AI and analytics at Capgemini, firms may gain a competitive edge by developing their own platform for sentiment analytics: Among huge corporations, this is the trend we're witnessing." In other words, they'll lose their competitive edge if they use the same generic tool that their rivals do. Application Programming Interfaces (APIs) are often used by firms establishing their own platforms to include sentiment analysis capabilities (APIs).

These features are available from all of the main cloud service providers. In the end, it comes down to the quality of these APIs. Simion reminds out that "if you have a specialised product, they generally won't benefit them." As for the price, he adds: "Each and every API request costs money. You need to make sure that calling these APIs is financially viable. "However, for small and medium-sized firms, this might be a viable option." The most common usage of sentiment analysis in the business world is in customer service, such as in contact centres. After a new product is released, organisations may look into the issue.

More and more video training data may be gathered as more and more web-based video conversations are done in customer service. The same technology that is used to analyse consumer sentiment may be used to analyse staff sentiment. A chatbot used by Genpact's Global Leader of Analytics Amaresh Tripathy reveals: "We utilise an artificial intelligence tool, a chatbot, for discussions. A chatbot, rather than the HR department, calls workers to see whether they need assistance. Companies might benefit from sentiment research when it comes to their brand image. By studying how consumers perceive a product, the company may tailor its marketing efforts toward those customers.

Conclusion

Twitter does not rely just on the frequency with which a word appears in a tweet to determine the most popular trending topics. If a significant quantity of data must be sorted using a mix of criteria with different weights, algorithms are used. WikiLeaks was often mentioned in tweets, yet its subtle dynamic had little effect on the algorithm's trend identification. The typical problem does not remain in the top 10 trending topics for more than 20 to 40 minutes. This is a variable measurement of how long individuals spend on the most popular topics.

Long-term hot topics have an abundance of "unique authors," which may have a factor in this instance. Each day, more than 50 million tweets are sent in English. Twitter users who retweet a term in an effort to spread it may be regarded both the trend's creator and the starting point for the critical mass of future mentions. Negative-negative sentiment analysis is the hedonometer's most common kind of analysis. Certain systems use a percentage or assume three values (positive, negative and neutral).

Sadness, rage, and joy are among the most prominent emotions. IBM Watson Discovery and Micro Focus IDOL provide customer service and marketing products sentiment analysis options. By creating their own platform for sentiment analytics, businesses may acquire a competitive advantage. Commonly, open source libraries serve as the basis for self-created applications. Amaresh Tripathy of Genpact said, "We use an artificial intelligence tool, a chatbot, for talks."

References

 de Oliveira Santini, F., Ladeira, W. J., Pinto, D. C., Herter, M. M., Sampaio, C. H., & Babin, B. J. (2020). Customer engagement in social media: a framework and meta-analysis. Journal of the Academy of Marketing Science, 48. https://doi.org/10.1007/s11747-020-00731-5

Makita, M., Mas-Bleda, A., Stuart, E., & Thelwall, M. (2019). Ageing, old age and older adults: a social media analysis of dominant topics and discourses. Ageing and Society, 41(2), 1–26. https://doi.org/10.1017/s0144686x19001016

Yue, L., Chen, W., Li, X., Zuo, W., & Yin, M. (2018). A survey of sentiment analysis in social media. Knowledge and Information Systems, 60(2), 617–663. https://doi.org/10.1007/s10115-018-1236-4