Come back after SxSW for a more detailed portfolio and navigation.
Deliverables
One way to finance our entry into a new space was to seek out grants. These grants could help us fund experiments with new segments and new product designs, preparing us to launch with well-tested products and methods. The manual search, however, would prove too time-consuming. Could there be a way to automate this type of search?
I broke this problem into a two-step process. At first, I used data mining to create a search strategy. I tested and tweaked the search parameters until the searches returned satisfactory results. I then automated the search so it would continue monitoring opportunities while we shifted our attention towards other activities.
Web-scraping and quantitive text analysis allowed me to identify the most common terms in the market we were looking into. I created, tested and tweaked regular expressions using these terms, to find the optimal search expressions.
I also tapped into interviews and calls with experts from our network, as well as searches in specialized data bases to identify terms. These became the basis of the keyword-based automated search tool I built.
I recorded my early findings in spreadsheets, allowing me to keep track of terms and categorizing them as I collected them. This became a reference pool for early trials around regular expressions.
Once the regular expressions had been finalized, I worked on automating these queries. I determined the frequency with which to conduct these searches, as well as the way results would be documented, archived, reviewed and shared with the executive team.
This allowed us for constant, weekly monitoring while freeing up resources for other tasks.
Strategist
Data Analyst
Project Manager