PHD Research Proposal in Machine learning Techniques for Social Media Analytics

With the increasing growth of the blogosphere has enabled consumers to persuade public perception. As a consequence, the availability of valuable insights also extensive, such insights helps marketing organizations to gather the people opinion to enhance the performance. On the other hand, the extraction of potential insights from the enormous data is a more challenging task. This kind of difficulties overcome by social media analytics [1]. Social media analytics concerning about the process of gathering and interpreting the data from the online social media to perform market research, and strategic business decisions. Social media analytics attempts to harvest the insights of the potential consumer from social media through the machine learning techniques that derive the hidden insights from the massive unstructured online social media data [2]. With this technique, the sentiment, emotion, audience demographics, topics, the opinion of the customer can be extracted to retrieve the most relevant data to enable insightful decision-making.
It has been broadly used in the recommender system, sentiment analysis, text mining, business approach and marketing [3].A key aspect of social networks is richer in information that brings the challenges and opportunities in the aspect of data mining and knowledge discovery [4]. The most challenging segments of social media analytics are the data coverage in which the tracking and collecting the rich content qualitative data is difficult. Unfortunately, not entire users of social media networks are truthful to spread information that induces the complexity while extracting the trustworthy information. The issue of determining the information spreaders requires the external force, yet it still leaves the research challenges. In the social networks, diverge content is exhibited widely different language style that makes it difficult to interpret the sentiment and context.
It results in significant complexity in the sentiment analysis of the free form text for decision making. Furthermore, the automated sentiment analysis still faces issues while dealing with the noise and free form text of the users in social media.
The rich user interaction and user-generated content on social media networks make difficulty in storage and processing power. The collaborative environment complicates the process of collecting the “representative” data about the entity from plenty of sources. An unprecedented scale of users and their unstructured data in the social media networks complicate the extraction of intelligence from that data that demands the expertise and scientific tool to manage the vast interconnected system. Furthermore, the data from several social media networks is noisy that complicated the identification of potential insights. Also, it makes difficult information diffusion in social media networks.

Reference:

  • [1] Stieglitz, Stefan, Linh Dang-Xuan, Axel Bruns, and Christoph Neuberger, “Social media analytics”, Business & Information Systems Engineering, Vol.6, No.2, pp.89-96, 2014.

  • [2] KORDA, PETAR, and PAVLE VIDANOVIC, “Machine learning techniques for social media analysis”, 2017.

  • [3] Ghani, Norjihan Abdul, Suraya Hamid, Ibrahim Abaker Targio Hashem, and Ejaz Ahmed, “Social media big data analytics: A survey”, Computers in Human Behavior, 2018.

  • [4] Sapountzi, Androniki, and Kostas E. Psannis, “Social networking data analysis tools & challenges”, Future Generation Computer Systems, Vol.86, pp.893-913, 2018.

Leave Comment

Your email address will not be published. Required fields are marked *

clear formSubmit