Author: Philip Godfrey
What is OCI Language?
OCI Language is a cloud-based AI service provided by Oracle for performing sophisticated text analysis at scale.
What does it do?
Text Analysis is a broad field and is often time-consuming for business to manually analyse large volumes of text data. This is why there are multiple offerings from Oracle Language services to help with this, utilizing Natural Language Processing (NLP) to derive insights.
We have summarised some of the key features of OCI Language below, but there is a full list of service offerings later in the blog.
- Language Detection – identify and translate more than 75 languages
within written text. - Entity Recognition – Automatically recognises specific entities such as names of individuals, places, products, and organisations.
- Sentiment analysis – how positive or negative the tone of the text.
Why should your business use Language services?
- Gain Insights: Product perception and decision intelligence
- Increase Revenues: customer insights, predictive data analytics.
- Reduce Costs: Automate manual processes with classification, entity recognition and translation.
- Increase Compliance: Automatically detect protected information.
What are the key benefits of using OCI language?
- Meet your business needs with the ability to train custom models, to meet your own business needs.
- Faster development with off-the-shelf AI models for common use cases.
- Helps focusing on business outcomes rather than data science.
- Pay as you go, no upfront costs.
What are some examples of where these might be used in business?
These are some broad examples but demonstrate how these language models can be applied to multiple industries and business sectors.
Customer support
- Route tickets to the appropriate team based on topic or language.
- Assist global support centres with translated text.
- Generate insights such as product name, dealer, region, etc.
Marketing
- Track customers sentiment toward your brand vs. your competitors.
- Identify frequently mentioned phrases trends in social media and news.
- Translate feedback or survey results for generate insights.
Manufacturing
- Automate business process by extracting key information such as product codes, product issues in warranty claims.
- Find engineering requirements in documents and RFPs.
How does it work?
|
Language service Capabilities: Pretrained models
|
Definition
|
|---|---|
|
Language detection
|
Identifies the dominant language in a text document
|
|
Named entity recognition
|
Identifies specific entities such as names, locations, dates, and times
|
|
Sentiment analysis
|
Understanding emotion expressed in textual data and returning and confidence score (sentiment could be positive, negative or neutral)
|
|
PII identification & de-identification
|
Personally Identifiable Information (PII) can be anonymised / de-identified using ML algorithms
|
|
Key phrase extraction
|
Extracting key phrases from a document, to understand what topic is being discussed
|
|
Text classification
|
Provide a category to classify a document, from a set of pre-defined categories
|
As OCI Language consists of multiple models, which include both pre-trained and custom models, we wanted to show at a high-level how OCI Language service is applied to your infrastructure.
Interested to know more?
We’d be more than happy to discuss your requirements or questions, please get in touch here.
Contact us today to arrange an assessment or email:
Philip Godfrey
Principal Data Analytics & AI Consultant at Vertice
