{"id":25163,"date":"2022-01-14T16:54:54","date_gmt":"2022-01-14T16:54:54","guid":{"rendered":"https:\/\/www.rightsdirect.com\/?post_type=blog_post&p=25163"},"modified":"2022-01-14T17:01:22","modified_gmt":"2022-01-14T17:01:22","slug":"semantischen-suche","status":"publish","type":"blog_post","link":"https:\/\/www.rightsdirect.com\/de\/blog\/semantischen-suche\/","title":{"rendered":"Wie sieht die Zukunft der semantischen Suche in der Wissenschaft aus?"},"content":{"rendered":"\n

Jeder kennt Suchmaschinen wie Google oder Bing. Die wissenschaftliche Suche ist jedoch ein spezieller Bereich der Suchindustrie. Wenn die Literatur Ihres Fachgebiets Aspekte wie kontrolliertes Vokabular (m\u00f6glicherweise einschlie\u00dflich Taxonomien oder Ontologien) und andere fachspezifische Metadaten umfasst, sind Sie als Forscher und Autor wahrscheinlich mit einem semantischen, auf Ihre Disziplin zugeschnittenen Suchwerkzeug am besten bedient.<\/h3>\n\n\n\n

Unser Kollege Dave Davis hat einen Blick auf die Zukunft der semantischen Suche geworfen und dazu einen Blogbeitrag ver\u00f6ffentlicht, den wir nachfolgend im Original ver\u00f6ffentlichen.<\/h3>\n\n\n\n

<\/p>\n\n\n\n

Many years ago, when I was learning about \u201ccommand languages\u201d used in online search tools such as DIALOG and BRS, I recall asking about certain strange (to me) groups of characters that made no natural-language sense to me, the first time I saw them. They appeared in only one of the databases we searched. My professor, a search language guru, explained: \u201cThose are reserved codes used in the Oil & Gas industries. And nowhere else,\u201d she intoned with the voice of wisdom and experience.<\/p>\n\n\n\n

Since the advent of generalized search engines \u2013 such as Google<\/a>, but in fact there were several good ones around a few years before<\/a> that \u2013 everyone knows (or think they know) about the basics of searching for information via their computer and a browser.<\/p>\n\n\n\n

Scientific search, by contrast, is a specialized corner of the search industry. When the literature of your subject discipline entails such aspects as a controlled vocabulary (possibly including taxonomies and ontologies) and other discipline-specific metadata, as a researcher and author you are probably better served by a semantic domain-aware search tool.<\/p>\n\n\n\n

<\/p>\n\n\n\n

For example: A screenshot from NLM\u2019s PubChem<\/a> search utility:<\/p>\n\n\n\n

\"screenshot<\/p>\n\n\n\n

<\/p>\n\n\n\n

Note: Although I used an example from chemistry, one could as easily look to Physics or Geology. There are probably as many discipline-specific tools as there are scientific domains.<\/p>\n\n\n\n

It follows that a search tool optimized for organic chemistry is not likely to work well at all when applied to questions in quantum astrophysics. They are not, as the idiom has it, \u201ctalking the same language.\u201d One approach, however, that is useful across disciplines is that known as \u201csemantic search.\u201d<\/a> Semantic search basically describes a cluster of techniques<\/a> that enable the algorithms of a search utility to probe a dataset for concept correlations (words and phrases traveling together) and, from these, to draw reliable inferences about \u2018hidden\u2019 relationships lurking in the text. It is an interesting angle, I think, because it goes beyond the simple \u201ctype in a bunch of keywords and hope for the best<\/em>\u201d tactic that most of us use, and it applies those semantic techniques to get more relevant results, more quickly, when the datasets include predefined scientific concepts. It\u2019s something I have been interested in (at a distance) for a while.<\/p>\n\n\n\n

Some of these thoughts came to mind again recently when I read an announcement from my employer (CCC) and its partner SciBite which heralded the incorporation of semantic search into our product RightFind Navigate<\/a>. Initially developed for use in the biomedical domain, this approach has opened to me possibilities\u2013 for better-targeted concept searching and e-discovery across scientific domains \u2013 that seem huge.<\/p>\n\n\n\n

While I\u2019m not in contact with my old prof. anymore, I wish I were so we could sit down over coffee and discuss the brave new world these new tools may open.<\/p>\n\n\n\n

Read more about Semantic Search & Enrichment:<\/strong><\/p>\n\n\n\n