Text Analytics

Negative Earnings Call Tone? Go Short (but not always)

Posts like “Facebook’s Q4: Conference Call Tone Matters More Than Results” in the financial press suggest that earnings call tone is important. And, invariably, the tone of the call does come up during the post call commentary and analysis. Were executives overly positive in their comments? Did they mean what they said? Did analysts’ questions…

Read More

Context, Then Concepts, Words Last

The “five forces of context” (mobile, social media, data, sensors and location) have be called the future of computing. Why? Because they may finally give computers the ability to understand “your context”. Analysts under time and deadline pressure need to know that the information distilled by an AI solution is relevant to their context and…

Read More

Good Concept Detection Requires an “Almost Engine”

Is “almost” good enough?  In terms of concept detection, the answer is most certainly “yes”.  In another guest post by Tom Marsh, CTO at Boulder Equity Analytics, Tom argues that textual analysis using BEA’s AI software allows analysts to efficiently cull through mounds of documents to eliminate the noise.  What is left are “scored” paragraphs…

Read More

Concepts are Key, Not Words

Some form of textual analysis has become a standard feature among services that offer summaries of large volumes of documents.  Natural Language Processing (NLP), deep learning and neural nets are buzz words we often hear.  But when you look under the hood, most of the functionality is based on keywords, word counts and rigid taxonomies. …

Read More