Skip to content

Commit 804f429

Browse files
committed
Add news markdown files in all languages (es, ca, en, eu)
1 parent 0605426 commit 804f429

3 files changed

Lines changed: 381 additions & 0 deletions

File tree

static/news/news_ca.md

Lines changed: 127 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,127 @@
1+
# LAMB - Learning Assistants Manager and Builder
2+
3+
**Plataforma web per projectar, entrenar i publicar assistents d'aprenentatge basats en IA sense escriure codi**
4+
5+
---
6+
7+
## Què és LAMB?
8+
9+
**LAMB** és una plataforma web de codi obert que et permet **dissenyar, entrenar i publicar assistents d'aprenentatge basats en IA** de forma visual i intuïtiva. Funciona com un "constructor de chatbots docents" que combina models de llenguatge (GPT-4, Mistral, models locals) amb els teus propis materials educatius.
10+
11+
**LAMB** és un projecte de codi obert desenvolupat per Marc Alier i Juanan Pereira, professors i investigadors de la Universitat Politècnica de Catalunya (UPC) i Universidad del País Vasco (UPV/EHU).
12+
13+
- **Website:** [http://www.lamb-project.org](http://www.lamb-project.org)
14+
- **GitHub:** [https://github.com/Lamb-Project/lamb](https://github.com/Lamb-Project/lamb)
15+
- **Llicència:** GPL v3
16+
- **Manifest:** [Safe AI Education](https://manifesto.safeaieducation.org)
17+
18+
---
19+
20+
## Característiques Principals
21+
22+
### 🎓 Tutors Temàtics Especialitzats
23+
Dissenya assistents que només responen sobre la matèria que triïs, mantenint-se sempre en el context educatiu apropiat.
24+
25+
### 📚 Ingestió Intel·ligent de Coneixement
26+
Puja documents (PDF, Word, Markdown) i LAMB els processa automàticament amb un **model de dades flexible** que:
27+
- Extreu i estructura el contingut preservant context i relacions
28+
- Crea embeddings semàntics optimitzats per a cerca educativa
29+
- Permet metadades personalitzades per a cada document
30+
- S'adapta a diferents formats i estructures de contingut
31+
- Alimenta el model mitjançant RAG (Retrieval Augmented Generation)
32+
33+
### 🔍 Proves i Depuració Avançades
34+
Mode "debug" que mostra el prompt complet per entendre exactament què s'envia al model, facilitant l'optimització de respostes.
35+
36+
### 🎯 Integració LTI amb Moodle
37+
Publica l'assistent com a eina externa LTI i incrusta'l al teu Moodle en un parell de clics.
38+
39+
### 🔒 Privacitat Garantida
40+
Els estudiants interactuen dins de LAMB; les seves dades no es comparteixen amb proveïdors externs de models d'IA.
41+
42+
---
43+
44+
## Per a qui és LAMB?
45+
46+
- **📖 Docents i formadors** que necessiten un ajudant virtual centrat en la seva temàtica específica
47+
- **🏫 Centres educatius** que usen Moodle o altres LMS i requereixen integrar IA sense exposar dades estudiantils
48+
- **💡 Equips d'innovació** que experimenten amb diferents LLM i necessiten un panell unificat de gestió
49+
50+
---
51+
52+
## Funcionalitats Clau
53+
54+
### Assistents Il·limitats
55+
Cadascun amb les seves pròpies instruccions, to i límits personalitzats.
56+
57+
### Bases de Coneixement Flexibles
58+
- **Model de dades adaptable**: arquitectura flexible que permet diferents tipus de contingut i estructures
59+
- Suport per PDF, DOCX, Markdown (més formats pròximament)
60+
- Bases públiques o privades segons necessitats
61+
- Sistema d'embeddings vectorials per a cerca semàntica
62+
- Connectors en desenvolupament per a fonts externes (Google Drive, YouTube, APIs)
63+
64+
### Múltiples Models d'IA
65+
- OpenAI GPT-4o
66+
- Mistral
67+
- Models locals
68+
- Canvi de model en un clic
69+
70+
### Cites Automàtiques
71+
L'assistent proporciona respostes amb referències als documents font utilitzats.
72+
73+
### Portabilitat Total
74+
- Exportar/importar assistents en format JSON
75+
- Fàcil versionat i compartició
76+
77+
### Interfície Multilingüe
78+
Català, castellà, anglès i euskera inclosos de sèrie.
79+
80+
### Control d'Accés Robust
81+
- Claus secretes per a registre
82+
- Bases privades per evitar usos no autoritzats
83+
84+
### Ecosistema en Creixement
85+
- **Arquitectura modular i extensible**: dissenyada per incorporar noves funcionalitats sense afectar el nucli
86+
- Plugins d'ingestió personalitzables per a diferents fonts de dades
87+
- API oberta per a integracions amb tercers
88+
- Actualització contínua sense dependència d'un sol proveïdor d'IA
89+
- Model de dades flexible que evoluciona amb les necessitats educatives
90+
91+
---
92+
93+
## En poques paraules
94+
95+
**LAMB et dóna el control total per construir un "ChatGPT especialitzat" en la teva assignatura, connectar-lo al teu Moodle i mantenir les dades del teu alumnat completament segures.**
96+
97+
---
98+
99+
## 🎓 Recerca i Fonament Acadèmic
100+
101+
**LAMB** està construït sobre recerca acadèmica sòlida i s'adhereix al **[Manifest d'IA Segura en Educació](https://manifesto.safeaieducation.org)** - un marc integral per al desplegament ètic, segur i educativament alineat de la IA.
102+
103+
### 📚 Publicació Acadèmica
104+
105+
Si fas servir LAMB en la teva recerca, si us plau cita el nostre treball:
106+
107+
**"LAMB: An open-source software framework to create artificial intelligence assistants deployed and integrated into learning management systems"**
108+
109+
- **Autors:** Marc Alier, Juanan Pereira, Francisco José García-Peñalvo, Maria Jose Casañ, Jose Cabré
110+
- **Revista:** Computer Standards & Interfaces, Volum 92, Març 2025
111+
- **DOI:** [10.1016/j.csi.2024.103940](https://doi.org/10.1016/j.csi.2024.103940)
112+
113+
### 🏛️ Socis Acadèmics
114+
115+
- **Universidad del País Vasco (UPV/EHU)** - Institució de recerca i soci de desenvolupament
116+
- **Universitat Politècnica de Catalunya (UPC)** - Institució de recerca i soci de desenvolupament
117+
- Facultat d'Informàtica de Barcelona
118+
- Institut de Ciències de l'Educació - ICE
119+
- Departament d'Enginyeria de Serveis i Sistemes d'Informació (ESSI)
120+
121+
### 🙏 Agraïments
122+
123+
Agraïments especials al **Projecte Open WebUI**, **Projecte Tsugi** (Dr. Chuck Severance), la comunitat de **Conferència TEEM**, i **Tknika** Centre de Recerca Aplicada de FP Basca pel seu suport i col·laboració.
124+
125+
---
126+
127+
*Última actualització: Desembre 2025*

static/news/news_en.md

Lines changed: 127 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,127 @@
1+
# LAMB - Learning Assistants Manager and Builder
2+
3+
**Web platform to design, train and publish AI-based learning assistants without writing code**
4+
5+
---
6+
7+
## What is LAMB?
8+
9+
**LAMB** is an open source web platform that allows you to **design, train and publish AI-based learning assistants** in a visual and intuitive way. It works as a "teaching chatbot builder" that combines language models (GPT-4, Mistral, local models) with your own educational materials.
10+
11+
**LAMB** is an open source project developed by Marc Alier and Juanan Pereira, professors and researchers at the Universitat Politècnica de Catalunya (UPC) and Universidad del País Vasco (UPV/EHU).
12+
13+
- **Website:** [http://www.lamb-project.org](http://www.lamb-project.org)
14+
- **GitHub:** [https://github.com/Lamb-Project/lamb](https://github.com/Lamb-Project/lamb)
15+
- **License:** GPL v3
16+
- **Manifesto:** [Safe AI Education](https://manifesto.safeaieducation.org)
17+
18+
---
19+
20+
## Main Features
21+
22+
### 🎓 Specialized Subject Tutors
23+
Design assistants that only respond about the subject you choose, always staying within the appropriate educational context.
24+
25+
### 📚 Intelligent Knowledge Ingestion
26+
Upload documents (PDF, Word, Markdown) and LAMB processes them automatically with a **flexible data model** that:
27+
- Extracts and structures content while preserving context and relationships
28+
- Creates semantic embeddings optimized for educational search
29+
- Allows custom metadata for each document
30+
- Adapts to different formats and content structures
31+
- Feeds the model through RAG (Retrieval Augmented Generation)
32+
33+
### 🔍 Advanced Testing and Debugging
34+
"Debug" mode that shows the complete prompt to understand exactly what is sent to the model, facilitating response optimization.
35+
36+
### 🎯 LTI Integration with Moodle
37+
Publish the assistant as an external LTI tool and embed it in your Moodle in a couple of clicks.
38+
39+
### 🔒 Guaranteed Privacy
40+
Students interact within LAMB; their data is not shared with external AI model providers.
41+
42+
---
43+
44+
## Who is LAMB for?
45+
46+
- **📖 Teachers and trainers** who need a virtual assistant focused on their specific curriculum
47+
- **🏫 Educational centers** that use Moodle or other LMS and need to integrate AI without exposing student data
48+
- **💡 Innovation teams** experimenting with different LLMs and need a unified management panel
49+
50+
---
51+
52+
## Key Functionalities
53+
54+
### Unlimited Assistants
55+
Each with their own instructions, tone and personalized limits.
56+
57+
### Flexible Knowledge Bases
58+
- **Adaptable data model**: flexible architecture that allows different types of content and structures
59+
- Support for PDF, DOCX, Markdown (more formats coming soon)
60+
- Public or private bases according to needs
61+
- Vector embeddings system for semantic search
62+
- Connectors in development for external sources (Google Drive, YouTube, APIs)
63+
64+
### Multiple AI Models
65+
- OpenAI GPT-4o
66+
- Mistral
67+
- Local models
68+
- One-click model switching
69+
70+
### Automatic Citations
71+
The assistant provides responses with references to the source documents used.
72+
73+
### Total Portability
74+
- Export/import assistants in JSON format
75+
- Easy versioning and sharing
76+
77+
### Multilingual Interface
78+
Catalan, Spanish, English and Basque included as standard.
79+
80+
### Robust Access Control
81+
- Secret keys for registration
82+
- Private bases to prevent unauthorized use
83+
84+
### Growing Ecosystem
85+
- **Modular and extensible architecture**: designed to incorporate new features without affecting the core
86+
- Customizable ingestion plugins for different data sources
87+
- Open API for third-party integrations
88+
- Continuous updates without dependence on a single AI provider
89+
- Flexible data model that evolves with educational needs
90+
91+
---
92+
93+
## In short
94+
95+
**LAMB gives you total control to build a "specialized ChatGPT" for your subject, connect it to your Moodle and keep your students' data completely secure.**
96+
97+
---
98+
99+
## 🎓 Research & Academic Foundation
100+
101+
**LAMB** is built on solid academic research and adheres to the **[Safe AI in Education Manifesto](https://manifesto.safeaieducation.org)** - a comprehensive framework for ethical, secure, and educationally-aligned AI deployment.
102+
103+
### 📚 Academic Publication
104+
105+
If you use LAMB in your research, please cite our work:
106+
107+
**"LAMB: An open-source software framework to create artificial intelligence assistants deployed and integrated into learning management systems"**
108+
109+
- **Authors:** Marc Alier, Juanan Pereira, Francisco José García-Peñalvo, Maria Jose Casañ, Jose Cabré
110+
- **Journal:** Computer Standards & Interfaces, Volume 92, March 2025
111+
- **DOI:** [10.1016/j.csi.2024.103940](https://doi.org/10.1016/j.csi.2024.103940)
112+
113+
### 🏛️ Academic Partners
114+
115+
- **Universidad del País Vasco (UPV/EHU)** - Research institution and development partner
116+
- **Universitat Politècnica de Catalunya (UPC)** - Research institution and development partner
117+
- Barcelona School of Informatics
118+
- Institut de Ciències de l'Educació - ICE
119+
- Department of Service and Information System Engineering (ESSI)
120+
121+
### 🙏 Acknowledgments
122+
123+
Special thanks to the **Open WebUI Project**, **Tsugi Project** (Dr. Chuck Severance), **TEEM Conference** community, and **Tknika** Basque VET Applied Research Centre for their support and collaboration.
124+
125+
---
126+
127+
*Last updated: December 2025*

static/news/news_eu.md

Lines changed: 127 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,127 @@
1+
# LAMB - Learning Assistants Manager and Builder
2+
3+
**IAn oinarritutako ikaskuntza-laguntzaileak kodea idatzi gabe diseinatzeko, entrenatzeko eta argitaratzeko web plataforma**
4+
5+
---
6+
7+
## Zer da LAMB?
8+
9+
**LAMB** kode irekiko web-plataforma bat da, **IAn oinarritutako ikaskuntza-laguntzaileak diseinatzeko, entrenatzeko eta argitaratzeko** modu bisual eta intuitiboan. "Irakasle-chatbot eraikitzaile" gisa funtzionatzen du, hizkuntza-modeloak (GPT-4, Mistral, tokiko modeloak) zure hezkuntza-materialekin konbinatuz.
10+
11+
**LAMB** Marc Alier eta Juanan Pereira UPCko (Kataluniako Unibertsitate Politeknikoa) eta EHUko (Euskal Herriko Unibertsitatea) irakasle eta ikertzaileek garatutako kode irekiko proiektu bat da.
12+
13+
- **Website:** [http://www.lamb-project.org](http://www.lamb-project.org)
14+
- **GitHub:** [https://github.com/Lamb-Project/lamb](https://github.com/Lamb-Project/lamb)
15+
- **Lizentzia:** GPL v3
16+
- **Manifestua:** [Safe AI Education](https://manifesto.safeaieducation.org)
17+
18+
---
19+
20+
## Ezaugarri Nagusiak
21+
22+
### 🎓 Gai Berezidunen Tutorak
23+
Hautatu duzun gaiari buruz bakarrik erantzuten duten laguntzaileak diseinatu, beti hezkuntza-testuinguru egokian mantenduz.
24+
25+
### 📚 Ezagutza Jasotzea Adimentsua
26+
Kargatu dokumentuak (PDF, Word, Markdown) eta LAMBek automatikoki prozesatzen ditu **datu-eredu malgu** batekin:
27+
- Edukia ateratzen eta egituratzen du testuingurua eta erlazioak mantenduz
28+
- Hezkuntza-bilaketarako optimizatutako embedding semantikoak sortzen ditu
29+
- Dokumentu bakoitzerako metadatu pertsonalizatuak ahalbidetzen ditu
30+
- Formatu eta eduki-egitura desberdinetara egokitzen da
31+
- RAG (Retrieval Augmented Generation) bidez modeloa elikatzen du
32+
33+
### 🔍 Proba eta Arazketa Aurreratuak
34+
"Debug" modua, modelora bidaltzen den promenada osoa erakusten duena, erantzunen optimizazioa erraztuz.
35+
36+
### 🎯 Moodlerekin LTI Integrazioa
37+
Argitaratu laguntzailea kanpoko LTI tresna gisa eta txertatu zure Moodlen klik bi batzuetan.
38+
39+
### 🔒 Pribatutasuna Bermatua
40+
Ikasleek LAMBen barnean elkarreragiten dute; haien datuak ez dira IA modelo hornitzaile kanpokoekin partekatzen.
41+
42+
---
43+
44+
## Norentzat da LAMB?
45+
46+
- **📖 Irakasleak eta trebatzaileak** haien curriculum zehatzean zentratutako laguntzaile birtual bat behar dutenak
47+
- **🏫 Hezkuntza-zentroak** Moodle edo beste LMS batzuk erabiltzen dituztenak eta IA integratu behar dutenak ikasle-datuak agertu gabe
48+
- **💡 Berrikuntza-taldeak** LLM desberdinekin esperimentatzen dutenak eta kudeaketa-panel unifikatu bat behar dutenak
49+
50+
---
51+
52+
## Funtzionalitate Nagusiak
53+
54+
### Laguntzaile Mugagabeak
55+
Bakoitza bere jarraibide, doinua eta limite pertsonalizatuekin.
56+
57+
### Ezagutza-base Malguak
58+
- **Datu-eredu moldagarria**: eduki eta egitura mota desberdinak baimentzen dituen arkitektura malgua
59+
- PDF, DOCX, Markdown euskarria (formato gehiago laster)
60+
- Base publikoak edo pribatuak beharren arabera
61+
- Bilaketa semantikorako embedding bektorialen sistema
62+
- Kanpo-iturrietarako konektoreak garatzen (Google Drive, YouTube, APIs)
63+
64+
### IA Modelo Anitzak
65+
- OpenAI GPT-4o
66+
- Mistral
67+
- Tokiko modeloak
68+
- Klik batean modelo aldaketa
69+
70+
### Aipamen Automatikoak
71+
Laguntzaileak erantzunak ematen ditu erabilitako iturri-dokumentuetako erreferentziarekin.
72+
73+
### Eramangarritasun Osoa
74+
- Laguntzaileak JSON formatuan esportatu/inportatu
75+
- Bertsio-kontrola eta partekatzea erraza
76+
77+
### Interfaze Eleaniztuna
78+
Euskera, gaztelania, ingelesa eta katalana serie gisa.
79+
80+
### Sarbide Kontrol Sendoa
81+
- Gako sekretuak erregistroarentzat
82+
- Base pribatuak erabilera baimenik gabeak saihesteko
83+
84+
### Hazten ari den Ekosistema
85+
- **Arkitektura modular eta hedagarria**: funtzionalitate berriak nukleoa ukitu gabe txertatzeko diseinatua
86+
- Datu-iturri desberdinetarako ingesta-plugin pertsonalizagarriak
87+
- Hirugarrenekin integrazioetarako API irekia
88+
- Etengabeko eguneraketak IA hornitzaile bakar baten menpekotasunik gabe
89+
- Hezkuntza-beharrekin eboluzionatzen duen datu-eredu malgua
90+
91+
---
92+
93+
## Labur esanda
94+
95+
**LAMBek kontrol osoa ematen dizu zure gairako "ChatGPT espezializatu" bat eraikitzeko, zure Moodlera konektatzeko eta zure ikasleen datuak erabat seguru mantentzeko.**
96+
97+
---
98+
99+
## 🎓 Ikerketa eta Oinarri Akademikoa
100+
101+
**LAMB** ikerketa akademiko sendoan eraikita dago eta **[Hezkuntzako IA Seguruko Manifestua](https://manifesto.safeaieducation.org)**ri atxikitzen zaio - IAren hedapen etiko, seguru eta hezkuntza-lerrokatua lortzeko esparru integral bat.
102+
103+
### 📚 Argitalpen Akademikoa
104+
105+
LAMB zure ikerketetan erabiltzen baduzu, mesedez aipatu gure lana:
106+
107+
**"LAMB: An open-source software framework to create artificial intelligence assistants deployed and integrated into learning management systems"**
108+
109+
- **Egileak:** Marc Alier, Juanan Pereira, Francisco José García-Peñalvo, Maria Jose Casañ, Jose Cabré
110+
- **Aldizkaria:** Computer Standards & Interfaces, 92. bolumena, 2025ko martxoa
111+
- **DOI:** [10.1016/j.csi.2024.103940](https://doi.org/10.1016/j.csi.2024.103940)
112+
113+
### 🏛️ Bazkide Akademikoak
114+
115+
- **Euskal Herriko Unibertsitatea (EHU)** - Ikerketa erakundea eta garapen bazkidea
116+
- **Kataluniako Unibertsitate Politeknikoa (UPC)** - Ikerketa erakundea eta garapen bazkidea
117+
- Bartzelonako Informatika Fakultatea
118+
- Institut de Ciències de l'Educació - ICE
119+
- Zerbitzu eta Informazio Sistemen Ingeniaritza Saila (ESSI)
120+
121+
### 🙏 Eskerrak
122+
123+
Esker bereziak **Open WebUI Proiektuari**, **Tsugi Proiektuari** (Dr. Chuck Severance), **TEEM Konferentzia** komunitateari, eta **Tknika** Euskal LHko Ikerketa Aplikatuko Zentroari haien laguntzarengatik eta kolaborazioagatik.
124+
125+
---
126+
127+
*Azken eguneratzea: 2025eko abendua*

0 commit comments

Comments
 (0)