Forecasting and Management of Technology, Second Edition

5.3 Structuring the Search

Gathering forecasting data is an iterative process. At each step, a little more about the topic under investigation is uncovered. Sometimes the right approach is to press ahead, complete the original search for information, and then consider a follow-up. At other times, it is best to restart the process with a new query that more accurately captures the topic. Framing the search query is very important. An experienced technical librarian suggested the following general process:

• Describe the information you seek in general terms (e.g., list the subject areas).
• Nominate terms (words or phrases) that seem to capture that subject information.
• Translate the terms into search logic (i.e., Boolean phrasing).
• Determine the types of sources desired (e.g., patents applied for or granted; journal articles or conference presentations; books or popular articles).
• Consider which sources to search (which databases, websites, etc.).
• Try a small-scale search (e.g., the most recent year or so); assess the results; and refine.

The search for data can be based on a variety of different strategies including

• Substantive terms
• Names (people, institutions, regions, or countries)
• Indices or classifications
• Citations or hyperlinks

Consider the trade-offs between broad and narrow searches (Table 5.10). The choice depends on your topic and target tech mining uses. If an aim is to spot unusual, nonmainstream R&D, you want to capture items with the barest threads of association ...