Table 2.
Studies using automatic textual analysis methods at the marketing–finance interface.
| Data information |
Extracted text information |
||||||
|---|---|---|---|---|---|---|---|
| Data type (text producer) | Data source(s) | Volume | Classification (applied approach) |
||||
| Authors | Discipline | Sentiment classification | Content classification | Topic modeling | |||
| Tirunillai and Tellis (2012) | Marketing | Earned social media (consumers) | Consumer reviews on Amazon.com, Epinions.com, Yahoo! Shopping | Yes | Yes (MLa [naive Bayes, support vector machine]) | No | No |
| Solomon (2012) | Finance | News content (institutions) | Factiva | No | Yes (lexicon-based [Loughran & McDonald, 2011]) | No | No |
| Green and Jame (2013) | Finance | Firm/brand names (firm) | CSRP | No | No | Yes (company name fluency based on lexicon approach) | |
| Xiong and Bharadwaj (2013) | Marketing | News content (institutions) | Lydia/TextMap (text processing system for >500 newspapers) | No | Yes (lexicon-based [Godbole et al., 2007]) | No | No |
| Borah and Tellis (2016) | Marketing | Earned social media during product recalls (consumers) | Third-party provider (not named) offering data from car-related forums, blogs, and reviews | No | Yes (lexicon-based, from third-party provided) | No | No |
| Hsu and Lawrence (2016) | Marketing | Earned social media during product recalls (consumers) | Third-party provider Alterian offering data from social media mentions, blogs, tweets, posts, images, and conversations | Yes | Yes (lexicon-based, from-third party provider) | No | No |
| Kashmiri et al. (2017) | Marketing | Product-market similarity (firms) | Text-based Network Industry Classification (TNIC) database by Hoberg and Phillips (2016) | No | No | Yes (similarity of words used by firms i and j) | No |
| Bartov et al. (2018) | Accounting | Earned social media (consumers) | Twitter reseller GNIP | No | Yes (ML [naive Bayes] and lexicon-based [Loughran & McDonald, 2011, Harvard Psychosociological Dictionary]) | No | No |
| Colicev et al. (2018) | Marketing | Earned social media (consumers) | Third-party provider (not named) offering data from Facebook, Twitter, and YouTube | Yes | Yes (ML [naive Bayes]) | No | No |
| Panagopoulos et al. (2018) | Marketing | CEO external focus (firm) | 10-K reports | No | No | Yes (lexicon-based [Yadav et al., 2007]) | No |
| Product-market fluidity (firms) | Product-market fluidity database by Hoberg et al. (2014) | No | No | Yes (changing of product words by rivals that overlap with firm i's vocabulary) | No | ||
| Sorescu et al. (2018) | Marketing | Diffusion of innovation (society) | Google Books Ngram Viewer | Yes | No | No | No |
| Bhattacharya et al. (2019) | Marketing | Strategic orientation (firms) | 10-K reports | No | No | Yes (lexicon-based [Linguistics Inquiry and Word Count]) | No |
| Chen et al. (2019) | Finance | Fintech innovations (firms) | Patent filings | No | No | Yes (machine learning [support vector machines, neural networks) | No |
| Dotzel and Shankar (2019) | Marketing | Service innovation announcement quality (firms) | Lexis Nexis | No | No | No | Yes (ML [latent Dirichlet allocation]) (customer vs. technology vs. service emphasis) |
| Frennea et al. (2019) | Marketing | Consideration of receivables investments (firms) | 10-K reports | No | No | Yes (lexicon-based) | No |
| Green et al. (2019) | Finance | Employee satisfaction (employees) | Glassdoor | No | Yes (difference in number of words in Pros and Cons sections) | No | No |
a
ML = machine learning.