WONS 2027 22nd Wireless On-demand Network systems and Services Conference 25 - 27 January 2027, Cortina d'Ampezzo, Italy

Why We Built This

For years, WONS has brought together people who are genuinely curious about where wireless networking is headed and who care about each other's work along the way. We wanted to stay true to that spirit and keep welcoming new voices into the conversation, so we built an AI-assisted tool to help us look beyond the faces we already know. It reads the meaning behind the work WONS has published over the years, listens for that same curiosity echoed across the wider literature, and gently points us toward researchers whose interests feel like a natural fit, even people who have never crossed paths with WONS before. It's our way of reaching out a hand and helping the community keep growing, year after year.

Researcher Discovery

Every paper becomes a point in a space where nearby points mean similar research. Ask what sits near a WONS paper, and the answer points to the people behind those neighbours.

WONS PAPERS VECTOR SPACE · ALL PAPERS CANDIDATE AUTHORS the query set

How It Works

Six steps across three phases — build the library, map its meaning, then find the people.
ABuild the library
1

Collect the papers

Gather paper metadata — titles, abstracts, keywords, authors — from the wider literature, alongside the full set of WONS conference publications used as the reference.

wider literature + WONS → one paper collection

2

Structure the corpus

Organise the collection so that every paper knows its text, and every author is linked to the papers they wrote.

papers → papers ↔ authors linked

BMap the meaning
3

Embed each paper

Turn every paper into a single vector that captures what the work is about — so that closeness in meaning becomes closeness in space.

each paper's text → a point in vector space

Model SentenceTransformer · embeddinggemma-300m

4

Build the searchable space

Place every paper's point into one space that can find a paper's nearest neighbours in an instant, even across the whole corpus.

every paper point → a searchable space

Index USearch · HNSW graph

CFind the people
5

Match against WONS

For every WONS paper, pull the papers that sit closest to it in the space. Each match carries a similarity — how near the two pieces of work really are.

each WONS paper → its nearest papers

6

Rank the authors

Trace the matched papers back to their authors and score each by how close their work sits to WONS. Authors who already publish there are set aside.

nearest papers → a ranked shortlist

How an Author Earns a Rank

An author's score reflects how closely their work sits to the WONS papers they turned up near. The higher it sits, the more their research rhymes with the conference — without them being part of it yet.
› Closer work → higher rank.
› Already publishes at WONS? Set aside.
› Too few papers or matches? Filtered out.
› What's left, sorted high to low.