Online Information Conference

LANGUAGE AND SEMANTICS: WHAT CAN YOU DO FOR MY SEARCH ENGINE (AND FOR ME)?

Search engine technology was not designed to solve the challenges it is up against today: huge number of users and increasingly complex documents. It is one of the oldest software technologies we use nowadays.

For users, search process has shifted from being a service (provided by librarians) to a self-service, similar to ATMs. This generates frustration for users and pressure on search engines to improve performance and user-friendliness.

Semantics can help solve this problem. However, so far, results are falling short of expectations.

Current natural language technologies provide one of the key ingredients to integrate search and semantics:  the ability to extract content from text (text analytics).

Bitext is applying this approach to real-life projects in areas like citizen services and opinion analysis.

Learning points:

Search engines are not delivering at the level that users demand. Semantics can help search engines deliver the results users need

Semantics is a complex discipline in which language knowledge and world knowledge converge. It is seen as a holy grail for the search problem

Semantic search requires the compilation of costly resources, intense in manual work and hard to reuse (ontology population, text tagging). Language technologies (text analytics) are one of the key technologies that can automate these tasks

Speaker

• Antonio Valderrabanos, CEO, Bitext, Spain