Latest Jobs at Digital Divide Data (DDD Kenya)

Job Description

• Natural Language Inference (NLI) tasks: This is a task of determining whether the given sentences logically follow, unfollow (contradict), or are neutral to each other. The translator would be expected to translate pairs of sentences used for the NLI task.
• Causal Common sense reasoning tasks: Will include the process of capturing and understanding the causal dependencies between two sentences i.e.
• translating pairs of sentences to determine if two sentences ideally follow each other
• Master or learn the tools associated with translation work 
• Precisely follow the established translation process, e.g. always raise queries when there is a need to
• Learn the style guide and overall client’s requirements, e.g. stylistic preferences
• Improve based on the corrections made by the senior team members as well as the client reviewers
• Keep track of personal error statistics and work on the weak spots
• Take part in the internal training and the ones organized by the client
• Participate in the error arbitration process
• Take part in the internal quality meetings and the quality calls organized by the client for the whole team
• Constantly analyze own gaps and think of what assistance the senior members and the client can provide to help address those

Qualifications

• Bachelor's Degree or Diploma in Linguistics or a related course.
• Basic knowledge in computer systems and technologies

Preferred Skillset

• Must be fluent in at least two languages - the language of the source (e.g
• English/French) and target language. The target language should be one of the following: Amharic, Ewe, Hausa, Igbo, Kinyarwanda, Lingala, Luganda, Oromo, Sesotho, Shona, Swahili, Twi, Wolof, isiXhosa, Yoruba, or isiZulu.
• Must have prior experience doing translations.
• Understand the translation process, terminology, and localization techniques.
• Writing, and good communication skills

go to method of application »

Job Description

Intent Detection tasks: Given a particular intent e.g “change pin” in a banking domain, the annotator should come up with sentences that describe the intent or action “change pin”. Several intents would be provided in several domains like tourism, banking, health e.t.c.

Slot Filling Tasks: This includes extracting from the utterance the correct argument value for the slots of the intent. Basically identifying spans of words in an utterance that correspond to certain parameters (i.e., slots) of a user request/query.

Qualifications

Education Qualifications

• Bachelor's Degree or Diploma in Linguistics, Computational linguistics, or computer science.
• Basic knowledge of computer systems and technologies

Preferred Skillset

• Must be fluent in at least two languages -  language of the source (e.g English/French) and target language. The target language should be one of the following: Amharic, Ewe, Hausa, Igbo, Kinyarwanda, Lingala, Luganda, Oromo, Sesotho, Shona, Swahili, Twi, Wolof, isiXhosa, Yoruba, or isiZulu.
• Must have prior experience doing annotation work e.g named entity recognition or sentiment classification.
• Writing, and good communication skills