Methods for aggregating investor sentiment from social media

Abstract

Abstract Social media-based investor sentiment proxies provide a brand new solution to recognize and measure investor sentiment. Aggregating individual social media text sentiments into public sentiments through a specific aggregation method is a necessary part of constructing an investor sentiment index for social media. The choice or design of the aggregation method directly affects whether or not the researcher can capture the sentiment of the market. This study provides the first systematic review of mainstream methods for aggregating investor sentiment from social media. In addition, we systematically discuss some of the key issues of historical researchers in aggregating investor sentiment, such as neutral sentiment text, simple aggregation of directly aggregated text, etc. The findings suggest that the aggregation method used by researchers directly affects the reliability of investor sentiment indices. Therefore, scholars should carefully choose sentiment aggregation algorithms based on the combination of datasets and sentiment tracking tools and articulate their rationale. This study provides important references for behavioral finance, social media mining, and microinvestor sentiment metrics.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 16 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.