Text Analysis in the Workplace
The rise of data in modern technological ecosystem has opened up a diverse range of opportunities to improve and innovate. Its immense potential is highlighted in the context of organisational data due to the scale that changes can be implemented at to produce value. However, as the magnitude of data availability and relevant opportunities grow exponentially, as do the obstacles to leveraging data innovation for its benefits (Kayser et al., 2018). These can be summarised with the 9 Vs conceptualisation of big data in its velocity, volume, value, variability, veracity, visualisation, validity, volatility and variety (Owais & Hussein, 2016). As a result, the challenges of contending with this complexity often result in significant underutilisation of organisational data, which is largely unstructured in nature.
Comprising 80 to 90 percent of data (Madeshan, 2015), the profound variety imbalance heavily favours unstructured data use in organisational processes. Still, many face the challenge of having the appropriate human and infrastructural resources, especially for unstructured data processing, management and analytics. Even possessing the necessary resources, organisations must also contend with investigating questions that provide value, drawing insightful analysis and integrating data that can be vast in its complexity and volume; as well as producing results that are interpretable for decision makers through tools such as visualisation. These actions then should be surrounded good data management practices: strategies relating to data privacy, governance, legislation, accessibility, ownership (Sivarajah et al., 2017) and codes of conduct shaping how data can be used to develop organisation value. Despite these challenges, the benefits of unstructured data analysis are clear: economic value produced by more effective text mining and analytics could generate a further 300 billion USD for the American healthcare sector or 250 million USD for European public administration (McKinsey Global Institute, 2012; cited by Ittoo et al., 2016).
Text analysis itself encompasses a broad range of analytical tools used to explore text data. It includes older techniques such as labour intensive manually coded text; as well as more novel, computer-automated text mining; both developing analyses of text corpora (Antons et al., 2020). However, in the context of modern information technology and its capacity to produce significant volumes of data at increasingly rapid velocities, i.e., data explosion (Zhu et al., 2009), text mining provides greater value in sustainable value creation. Its techniques fall under two main categories of classification (supervised) and clustering (unsupervised) methods. The former seeks to assign categories for text derived objects, supervised by the input of pre-defined categories, whereas clustering methods group similar text objects without the supervision of input categories (Antons et al., 2020).
Text analysis can be presented in various ways to heighten interpretation of information. Summary statistics can quantifiy results of text mining into simpler, condensed forms that enable ease of performance evaluation and comparability, however interpretability may require some quantitative knowledge. This could be reduced with accompanying data visualisations to present text analyses graphically to those less acquainted with statistics. These tools enable representation of analyses in forms that are more consumable to decision makers of organisations, reducing vast volumes of data to information that is more easily incorporated into forming budgets, plans and strategy.
Applications of these techniques in combination with automated predictive machine learning has presented a wide range of innovative developments. Existing trends of these can be seen in natural language processing in online customer service chatbots and auto-complete functions on emails which interpret human language to generate something of value to users. In terms of organisational value, these data innovations enhance productivity by limiting time required for completion of mundane or labour-intensive tasks, in this example improving service efficacy of basic customer enquiries and constructing emails (Sharma, 2020). This value creation by internal improvement focuses on the organisation itself and the ways in which text analysis can drive efficiency of operational processes.
In contrast, strategic organisational value can also be attained with the aid of text analysis by focusing on external factors such as target markets and customers, external vendors and business environments. Examples of these include sentiment analysis of customer reviews and topic modelling service-related comments from technology or car technicians to extract insights to be used for future product development and service improvements. Externally focused text analysis could also be applied to supporting investment decisions: mined sentiment from news articles or social media feeds could be modelled to predict fluctuations in stock prices (Chen and Lazer, 2011). These techniques provide organisational value by finding opportunities for innovation and improvement, as well as supporting subsequent strategic decision making.
Text analysis has had a profound impact on the development of many data innovations, creating value for not only organisations and their internal stakeholders but also the end users of their services and products. It taps into under-utilised resource present in large quantities in many organisations to provide significant business value. More optimal use of text analysis can help inform operational, tactical and strategic planning by generating more interpretable, actionable insights to improve process effectiveness and efficiency, as well as allocations of time and resources. This creates greater opportunities for organisations to invest in more transformative functions to create novel ideas, products, services and innovations. Continued organisational application of text analysis techniques are likely to provide sustained value, not only for the organisation themselves, but also for the wider community innovation.
References
Antons, D., Grunwald, E., Cichy, P. & Salge, T. O. (2020) The application of text mining methods in innovation research: current state, evolution patterns, and development priorities, R&D Management, 50(3), pp. 329–351. https://doi.org/10.1111/radm.12408
Chen, R & Lazer, M. (2011) Analysis of Twitter Feeds for the Prediction of Stock Market Movement, Semantic Scholar, viewed 20 May 2021, <
Kayser, V., Nehrke & B., Zubovic, D. (2018) Data Science as an Innovation Challenge: From Big Data to Value Proposition, Technology Innovation Management Review, viewed 30 May 2021, <https://timreview.ca/article/1143>
Madeshan, N. (2015) An Exclusive Study on Unstructured Data Mining with Big Data, International Journal of Applied Engineering Resarch, 10(4), pp. 3875–3886.
Owais, S. S. & Hussein, N. S. (2016) Extract Five Categories CPIVW from the 9V’s Characteristics of the Big Data, IJACSA, 7(3). https://doi.org/10.14569/IJACSA.2016.070337
Sharma, A. (2020) Top 10 Applications of Natural Language Processing (NLP), Analytics Vidhya, viewed 30 May 2021, <https://www.analyticsvidhya.com/blog/2020/07/top-10-applications-of-natural-language-processing-nlp/>
Sivarajah, U., Kamal, M. M., Irani, Z. & Weerakkody, V. (2016) Critical analysis of Big Data challenges and analytical methods, Journal of Business Research, 70(2017), pp. 263–286.
https://doi.org/10.1016/j.jbusres.2016.08.001
Zhu, Y., Zhong, N. & Xiong, Y. (2009) Data Explosion, Data Nature and Dataology. In Zhong N., Li K., Lu S., Chen L., (eds). Brain Informatics. BI 2009. Lecture Notes in Computer Science, Springer, Berlin, Heidelberg. vol 5819. https://doi.org/10.1007/978-3-642-04954-5_25
