Attributes in Search Service

Searching for documents is an operation we perform daily, often multiple times a day.

However, every experienced internet user knows from experience that searching for a document can be a very difficult task indeed. In the past, keywords were used, and searches yielded too few results. Later, they transitioned to free text search, but we encountered the opposite problem: excessive flooding of results, which could deter users from searching altogether.

Therefore, they try to limit the search today. The technique: attributes (cataloging) that will form the basis for filtering documents. This is true on the external internet, and it is also necessary on the internal intranet. In organizations, the task is easier, as the basis for attributes is the organizational taxonomy, which serves as the internal organizational terminology dictionary. The taxonomy refers to a collection of everyday terms heard within the organization's walls, which an outsider would not understand, but employees are familiar with. Terms from the world of the organization's customers, technologies, and products.

Based on these terms, it is possible to define attributes and values that will form the basis for effective filtering of free search. Instead of searching, for example, for all documents containing the word "methodology," by using attributes, it is possible to narrow the search to methodology documents on the subject of "knowledge management" only. In other words, the reason for adding attributes to every document is to reduce search results quickly and effectively.

Methodologies:

There are three different methodologies for defining taxonomy and extracting attributes from it.

1. Online workshop: Based on an interactive workshop method. Building together with people, the list of attributes from everyday useful terms using yellow sticky notes and creativity.

2. Knowledge questionnaires: A method that builds the attributes and their values from analyzing knowledge questions we encounter in daily work.

3. One-on-one: A method that builds the attributes from the document tree on the network.

For more details on the various methods, please contact the monthly editors. Easy and pleasant searching!

---

Want to learn more about digital platforms?

Here are some articles you might find interesting:

• Federated Search System

• Blogs in Organizational Use

• One database for multiple target audiences