So, for the last eight years, we’ve built, re-built and are continuously fine-tuning the PitchBook Platform’s UX—putting special focus on understanding the way our users search for information, process, and analyze it.
Recently, we turned to data science technologies (think: machine learning and natural language processing) to make it easier for our users to run targeted searches and find exactly what they’re looking for.
Until this week, users running keyword-based searches in the PitchBook Platform unknowingly put invisible barriers on their search results—if they typed in the words “artificial intelligence,” their results were limited to only those entities that contained the words “artificial intelligence.”
A lot of databases operate like this, and see it as a value-adding feature that limits noise in complex search queries. We went along with it until we realized that this had the potential to cause our clients to miss out on opportunities—and that just wasn’t acceptable to us.
To address this problem, we explored a few options. Ultimately, we ended up developing a complex algorithm that recommends additional keywords, based on how often certain words appear together, and in the same context.
Now, when a PitchBook Platform user types the word “artificial intelligence” into their search, they’re given the option to select additional, related keywords such as “speech processing,” or “cognitive computing.”
I anticipate this feature will improve as we continue to expand our data and refine our techniques. Until then, I’d invite you to take it for a spin and let me know what you think, by logging in (if you’re a PitchBook Platform subscriber) or requesting a free trial.