TalkingPoints | Innovation

Innovation

TalkingPoints is the only organization focusing on a technology-based solution to family engagement for under-resourced, multilingual communities. TalkingPoints is fundamentally redesigning how educators and families work together to build meaningful relationships in support of their students.

We are constantly improving our product and our model to make sure we have the greatest impact on student success in under-resourced communities.

Our approach

TalkingPoints leverages technology to provide a low-cost, scalable and impactful approach to tackle the barriers to family engagement. By harnessing cutting-edge technology like machine learning algorithms and natural language processing, we believe we can build a stronger family engagement practices and home-school connection in support of student success.

Innovation in improving translation quality

Continuously providing high-quality, understandable translations of messages between English-speaking educators and non-English-speaking families is central to removing language barriers to home-school connections and therefore critical to our mission. We believe we’re the only organization investing in education-specific translation, and our status as a nonprofit organization enables us to do so.
We take a multi-pronged approach to continuously enhancing translation quality.

Public machine translation

We use existing machine translation platforms available  for baseline quality translation.

Human translators

Human translators support near-real-time translation  and contribute high-quality data to train  our own algorithms.

TalkingPoints machine translation

We train machine learning algorithms using our own data, which combines our conversation database  with human translations to make it education-domain specific. We have also developed education-specific glossaries to enhance translations.

Help me understand

Educators and families can request human translator review of any translated communication (English or non-English) for any messages that may not be immediately understandable.

Innovation for accessibility

As a nonprofit committed to ensuring that our platform is able to serve the most under-resourced communities, and be accessible to all, we are working on:

Innovation for building capacity

Based on the dual-capacity building framework for strong family-school partnerships, we believe in building capacity and knowledge, and empowering administrators, educators and families in a relevant, personalized and convenient way. What we’re working on to make it relevant and personalized:

Topic classification of conversations

Using natural language processing, we are investing in a topic categorization tool to understand the pulse and topic of conversations between educators and families. This will enable:

Data for personalization

We measure feedback and actions of families and educators to refine what in-app coaching content is helpful and useful to build capacity, and for whom, allowing us to eventually personalize the experience.

Research partnerships

We are committed to understanding, measuring and furthering our impact through research partnerships, and spurring innovation for the field.

Innovation

Our approach

Innovation in improving translation quality

Public machine translation

We use existing machine translation platforms available  for baseline quality translation.

Human translators

Human translators support near-real-time translation  and contribute high-quality data to train  our own algorithms.

TalkingPoints machine translation

We train machine learning algorithms using our own data, which combines our conversation database  with human translations to make it education-domain specific. We have also developed education-specific glossaries to enhance translations.

Help me understand

Educators and families can request human translator review of any translated communication (English or non-English) for any messages that may not be immediately understandable.

Innovation for accessibility

As a nonprofit committed to ensuring that our platform is able to serve the most under-resourced communities, and be accessible to all, we are working on:

Continued use of text messages

Text to speech

Multimedia captioning (upcoming)

Innovation for building capacity

Based on the dual-capacity building framework for strong family-school partnerships, we believe in building capacity and knowledge, and empowering administrators, educators and families in a relevant, personalized and convenient way. What we’re working on to make it relevant and personalized:

Topic classification of conversations

Using natural language processing, we are investing in a topic categorization tool to understand the pulse and topic of conversations between educators and families. This will enable:

Data for personalization

We measure feedback and actions of families and educators to refine what in-app coaching content is helpful and useful to build capacity, and for whom, allowing us to eventually personalize the experience.

Research partnerships

We are committed to understanding, measuring and furthering our impact through research partnerships, and spurring innovation for the field.

Innovation

Our approach

Innovation in improving translation quality

Public machine translation

We use existing machine translation platforms available for baseline quality translation.

Human translators

Human translators support near-real-time translation and contribute high-quality data to train our own algorithms.

TalkingPoints machine translation

We train machine learning algorithms using our own data, which combines our conversation database with human translations to make it education-domain specific. We have also developed education-specific glossaries to enhance translations.

Help me understand

Educators and families can request human translator review of any translated communication (English or non-English) for any messages that may not be immediately understandable.

Innovation for accessibility

As a nonprofit committed to ensuring that our platform is able to serve the most under-resourced communities, and be accessible to all, we are working on:

Continued use of text messages

Text to speech

Multimedia captioning (upcoming)

Innovation for building capacity

Based on the dual-capacity building framework for strong family-school partnerships, we believe in building capacity and knowledge, and empowering administrators, educators and families in a relevant, personalized and convenient way. What we’re working on to make it relevant and personalized:

Topic classification of conversations

Using natural language processing, we are investing in a topic categorization tool to understand the pulse and topic of conversations between educators and families. This will enable:

Data for personalization

We measure feedback and actions of families and educators to refine what in-app coaching content is helpful and useful to build capacity, and for whom, allowing us to eventually personalize the experience.

Research partnerships

We are committed to understanding, measuring and furthering our impact through research partnerships, and spurring innovation for the field.

We use existing machine translation platforms available  for baseline quality translation.

Human translators support near-real-time translation  and contribute high-quality data to train  our own algorithms.

We train machine learning algorithms using our own data, which combines our conversation database  with human translations to make it education-domain specific. We have also developed education-specific glossaries to enhance translations.