Search & Categorization for Knowledge Retrieval
One of the most important aspects in sharing and retrieving knowledge and information using an IT system is the ability to find what you are looking for. The more data, information, and knowledge that is stored in our computers, in databases, across networks, in the cloud, etc. the more difficult it is to locate what we need, quickly and effectively.
XU et al (2011) indicated that knowledge retrieval is one of the biggest challenges today, and it is something that many IT systems try to address. There are several ways an IT system can contribute towards information and knowledge retrieval, including powerful search functions with filtering options and intelligent search tools.
Information systems can thus have any number of features. These may include searching across different types of media, query assistance, ranking, filtering, support of different indexation methods, search by keyword or search by relevance, supporting different languages, etc. Techniques used in knowledge and information retrieval range from simple keyword indexing/searching to advanced algorithms and neural models.
Babu et al (2012) identify several general steps in the information retrieval process:
- Indexing: Here the indexer must “capture” what the document is about. This may be done automatically through sophisticated processes designed to extract key information – something that spans beyond simple text and can include images, sound, etc. – or manually by the user (which may include user-specified keywords, descriptions, abstracts, etc.).
- Query formulation: “The query description is transformed, manually or automatically, into a formal query representation…” “…that combines features that predict a document’s usefulness. The query expresses the information need in terms of the system’s conceptual schema, ready to be matched with document representations” Babu et al (2012)
- Selection: This is where the results are presented to the user – with some form of ranking, grouping, etc. The user selects the results they need and may be involved in improving the search system. Relevance feedback, for example, is considered to be an important tool to improve selection. For instance, Hofmann et al (2015) describe the different metrics used to evaluate online user relevance, grouping them as document-level, result-list-level, and session-level metrics.
Information and knowledge retrieval systems can thus be independent systems aimed at better searching through knowledge and information repositories (e.g. search engines, specialized search software, digital libraries, etc.) or they can be built into intranet systems, document retrieval systems, content management systems, etc. They can employ a multitude of features, depending on the types of searches they are meant to perform, so as to provide the user with the most efficient and effective searches of knowledge and information repositories.