In Finland, the analysis of large volumes of data is largely based on the analysis of structural and numerical data, as the automatic processing of the Finnish language is exceptionally challenging. Since summer 2018, Lingsoft's language analysis technology has been used to help IBM Watson Explorer understand Finnish and Swedish. This partnership will open new doors for Finnish-language text analysis.
IBM Watson Explorer is capable of understanding complex questions, even those posed in spoken language, as well as providing knowledge-based answers to them. It can be used to analyse large volumes of text as well as classify and find desired phenomena in various types of text, from patient records to customer feedback.
IBM Watson Explorer's analysis dives deep into a language. The Finnish language's wealth of conjugations and almost endless possibilities for forming new words make it more challenging to process than other languages. For example, adjectives only have a basic form in English, while in Finnish adjectives can theoretically take 6,000 different inflected forms. Language technologies developed for English are usually not suitable for the machine analysis of Finnish materials.
In order for its solutions to tackle the challenges posed by different language regions, IBM networks with specialists such as Lingsoft. "It's truly gratifying to gather the top minds from different fields around the same table. Together with Lingsoft and other companies in our ecosystem, we can generate completely unique value for our Finnish clients − something none of us could achieve alone," explains IBM Channel Manager Antti Haapasalo.
Lingsoft has developed its Finnish and Swedish language analysis solution systematically since 1986. The analysis is based on language-independent technologies that are used to dive deeper into a language's structure. In addition to language structure analysis, Lingsoft's technology can reach the semantic level of a language − its meaning. This makes it possible to recognise different classifications, such as names, expressions of time, emotions or even symptoms of diseases. IBM Watson Explorer's user interface allows different elements to be explored, visualised and demonstrated in an effective manner.