Organizational aspiration for social impact

Aspiration, especially in context of CSR reporting, can be defined as communication about social intentions and ideas that might not be fully reflected in organizational behavior, but they are attractive in messaging for society (Sobkowiak 2020).

Objectives

The objective of the research is to explore CSR reports what summarize the message about social impact the organization send to stakeholders.

the roles played by organizations in society

  • The essence of aspiration for social impact is to make stakeholders understand how an organization creates social benefits and how this is consistent with business goals (Stephan et al. 2016).
  • Social impact is conceptualized in the literature in such terms as social value, social performance, and social return on investment (Yates 2017; Nason et al. 2018), benefit-sharing, local community development, stakeholder engagement, sustainability, and long-term resilience (Vanclay 2020, Dubey et al. 2019; Ojansivu et al. 2020). So, it is important to check if the company is approaching this wording.
  • The research of social impact is focused on social performance (Kurniawati et al. 2017), social returns (Krlev et al. 2013), social return on investment (Hall et al. 2015) and social accounting (Nicholls 2009).

Researchers still just started to investigate the causes and that level of the organizational desire for social appreciation including the whole environment the organization is acting, that is why study of this field is still required.

Sustainability investigation diagram

Text mining – other practical implementations

  • Diagnosis supporting based on investigation summarization
  • Archive mining for information based on semantic similarity
  • Text, and speech understanding
  • Routine communication automation like chatbots and communication with machines
  • AI advertisement
  • Fake news, spam detections, etc. …
  • Information classification
  • Etc. an extensive list … of …

    Theories in focus

    • The resource-based theory suggests that a company creates relationships with business partners and society as the ecosystem with the aspiration and purpose of generating external resources (San and Salim 2017). It assumes a bilateral nature of relationships, where organization and stakeholders influence one another or others to obtain the best possible opportunities to compete on the market through resource extension.
    • Stakeholder theory examines the relationships with multiple stakeholders. It defines models for identifying the stakeholders that can be involved in relationship management and the extent to which they are viewed as channels for multiple impact. Stakeholder theory describes: (a) human actors co-operating in value creation; (b) values, norms, and ethics as mechanisms of success both within and among organizations; and (c) a worldview of business that reflects a higher state of consciousness about its short- and long-term role in and impact on society (Freeman et al. 2020).
    • Legitimacy theory explores the concept of legitimacy as the process where an organization is seeking social acceptance for its actions and aspirations (Suchman 1995, p. 574). Legitimacy management is a cultural process where an organization is trying to maintain and regain stakeholders to support its social orientation through communication with various audiences (Peng et al. 2021).
    LDA
    LDA by sectors

    Documents contribution to topics

    Contribution represents coefficient how many documents by probability summary contribute to the topic.

    document contribution to tocpics
    heatmap

    Aspiration in BERTopic and Top2Vec analysis

    • BERT is Bidirectional Encoder Representations from Transformers (BERT)
    • Grootendorst¹¹ (2022) and Angelov¹² (2020). BERT pretrains deep bidirectional representations from unlabeled text by jointly conditioning on both left and right context in all layers. BERT uses the Masked Language Model (MLM) to use the left and the right context during pre-training to create a deep bidirectional Transformers.
    • BERTopic and Top2Vec manufacture semantic embeddings from input documents. BERTopic leveraged BERT Sentence Transformers (SBERT) to manufacture high-quality, contextual word and sentence vector representations. Instead, Top2Vec used Doc2Vec to create jointly embedded word, document, and topic vectors. 
    • A dimensionality reduction is applied before clustering. UMAP (Uniform Manifold Approximation and Projection) (McInnes¹³ et al., 2018). Model is reducing embedding dimensionality with UMAP improves the clustering performances of K-Means and HDBSCAN in terms of accuracy and time (Allaoui¹⁴ et al., 2020).
    • Both BERTopic and Top2Vec originally leverage HDBSCAN (McInnes¹⁵ et al., 2017) as clustering algorithm.
    https://towardsdatascience.com/intuitive-explanation-of-bert-bidirectional-transformers-for-nlp-cdc1efc69c1e
    Bert topics according to sectors Ret, F&B and Tex

    Aspirations for social impact of the organization are utilized in CSR reporting in the form of the wording which company uses for messaging to the market. Specific wordings for segments and company are clearly visible and differentiate companies where they are pointing, their core business challenges.

    Aspirational talk can be detected on the algorithmic way to readable form. Its express characteristic of the company approach to social impact and the effort placing for business relationship management and stakeholder engagement. It is the essential part of the socially oriented activity. The wording for segments and company, its sentiments, and relations to selected patterns, as GRI standards or circular economy glossary we can process algorithmically. 

    Company aspiration for social impact expresses in the creation of shared value for society. It is clearly demonstrated in the wording detected from mission statement and strategy. It represents the strategy in a symbolic summary. The wordings for segments and company can be investigated as relations to mission statements and strategy formulation in an algorithmic way.  

    Similar Posts