Room type mapping
Room attributes separation
Hotel & vacation rental
Hospitality content mapping
Customer oriented
- Increasing discoverability
- Improve retention, conversion
- Enhancing overall SEO
- Providing a better UX
- Strengthening brand loyalty
- Fast forward selection
- Fine-tuning search for sales
Price optimization
- Compare margin per room range
- Increase booking sales
- Compare rates @ room avail.
- Snap more revenue on the fly
- Apply a dynamic markup model
- For high rewards & profitability.
- Lower L2B & enhance productivity
Artificial intelligence
- Room range rate prediction
- Personalization search engine
- pyTorch tensorflow OpenAi csv
- ABS automated onboarding
- Room attr granular compset
- Json & csv for data scientists
- Pre-trained & recursive models
a B2B2C tool
all in 1 solution
a powerful API
high automation
live instant speed
smart & accurate
Extensive compatibility
From a few to billions of rooms, we get you covered.
Parsing 100, 10000, 1 million availability calls to your suppliers every day ?
Processing 10 Billions rooms per month across 20 suppliers ?
Apply price optimization automation at search level.
Our solution instantly maps & sorts out suppliers' rooms.
Real time mapping
-
Dynamic price optimization
Cheapest per type & specs
Use room-matching dynamically to compare room rates returned by your suppliers.
We map & remove duplicates across the list, and group the similar ones together.
Use it to survey a preferred option search & catch the right time to book or to find the best opportunity at T-time.
Automatically filter who is having the cheapest offer.
Get the best rates across the hotel's rooms range.
Upgrading & repricing
When doing a booking rate survey by remapping the room across suppliers availabilities until check-in.
Goal is to check if the rate goes down & if so, process a cancel/rebook.
You can automate a recheck, depending on your best sellers, anticipated sells, bookings log, preferred season, cancel policy date.
Enhance your repricing model by also checking if an offer at the same hotel has better specs & attributes for the price.
"best value for the money" upgrading.
Dynamic markup
Applying a dynamic markup model is simpler than you think.
Needed: an OTA as supplier to get the BAR public rate & at least one B2B supplier or bed-bank (more is better).
Example:
Room BAR is $200. cheapest B2B is $120
Markup can be dynamically set at 50% or $80 for a win/win. $40 you, $40 customer. You control the margin intermediation.
Sort offers by profitability for high rewards.
Vertical B.I.
In some cases, a commission model is better, and in other scenarios, the gross wholesale model is.
Anticipate the best buying period. Right spot the best deal.
A collateral benefit is easing vertical distribution B.I. & parity check across actors.
Control margin split per room type & group either at hotel level or at a large scale.
Envision a global market profitability.
The cream of the crop
Zooming at the peak of the curve (golden triangle) at destination level search, there is a 2% chance to hit a +80% margin offer. Sometimes more. It depends on availabilities & the number of mapped suppliers.
Mapping hundreds of hotels at destination, you end up with 50 to 100k rooms mapped for profit.
More than 100 of them are constantly "rotating" available on a market, with a margin ranging from 80% to 200% (seen).
Snap the high $ rooms to push on your website.
Real-time mapping
To map & standardize your rooms within a minimal processing time.
For a TA, the cost of an additional 1 second lag for 1000 rooms (average number of rooms we see) is acceptable.
The productivity ratio is by far in favor of having rooms sorted out.
For B2C, OTA, you need to go faster. We have solutions for it.
Use dynamic mapping to compare rates & apply a dynamic markup model.
Use static mapping to create a cache & add personalization to your website.
Our solution works seamlessly with your system.
Dynamic explained
live dynamic mapping at search level
For any dynamic & static mapping jobs.
Our solution works seamlessly with your inventory & application.
Download a turnkey mapped database cache inventory.
What is NLP ? (Wikipedia)
Artificial intelligence
-
Attribute level Fintech
-
Next gen search engine
What to do with our datasets ?
-
-
-
-
-
-
-
-
-
Room-level search engine and mapper across inventories.
Room range rate forecasting across bookings log.
Revenue Management at selected attributes vs compSet.
Dynamic price optimization and markup at availability search.
Attribute-Based Sales ready mapping and datasets for A.I.
ChatGpt room recommendation csv upload ready.
Matched attributes parity check controller.
Price prediction at granular analytic level.
Repricing/upgrading option opportunity eval.
Ask us about advanced hospitality B.I. Apps, custom dataset, models, search engine, smart filtering, rooms UX show & sorting (consulting).
Upload, click & train your app.
(interface coming soon)
Upload your inventories and logs, select your application (use our presets), service provider and train your A.I.:
NLP
Room level search engine across suppliers
Pytorch
Room range rate forecasting across bookings log
ChatGpt
Room recommendation chatBot across inventory
NLP
Automated ABS oriented onboarding mapping
TensorFlow
RM at selected attributes versus compSet
NLP
Matched attributes parity check controller
A Machine Learning application uses a dataset with cleansed, qualified & standardized by attribute columns (see below).
Our mapping dataset is dedicated & formatted for hospitality granular analytics.
Unlimited mass mapping package for prediction:
Our API allows to map billions. Map your logs since covid. Special low cost package.
We provide consulting for trainers, transformers, tuning & various app models.
Labeling (annotation) can also be customized as other mapping aspects & preferences (custom solution).
Dataset preview:
20 attributes topics mapped in total, superior meaning labels for LLM at an additional cost, dataset available in english, mapping available in 5 languages. (see below features).
room_name | room_main | hotl_code | sppl_code | room_code | rate_plan | room_rate | meal_plan | attr_type | type_labl | attr_cteg | rang_labl | attr_view | view_labl | attr_amn1 | attr_bdrm | bdrm_labl | .... | fast_train | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
premium treehouse vue méditerranée 3 large bedroom | rental | 100258 | 1 | 1789654 | 65433 | 524 | ___ | treehouse | atypic | premium | luxe | mediterranean | sea | ___ | 3 | large, family | ... | [54.0000, 1.0000, 31.9160, 0.0000, 0.0000...] | |
suite highflr side beach vx | room | 100589 | 2 | 7896541 | 14785 | 240 | breakfast | suite | ___ | standard | ___ | side | beach | high floor | 1 | ___ | ... | [63.0000, 0.0000, 27.2498, 0.0000, 1.0000...] | |
... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
Using our datasets, you can offer an at-destination real-time search engine working at hotel, vacation rental & room levels with dynamic zoom filtering.
As hotel names, vacation rentals & rooms descriptions are mapped the same way, the dataset is standardized, normalized and unified across all levels, allowing precise personalized & by attribute search at destination across all markets.
you dynamically compare vacation rentals with rooms. You search rooms by hotel name or attributes.
Dynamic search mapping demo across 12 suppliers (b2b, GDS, bedbank, OTA).
Personalization brings more revenues
A live by room type & specs filtering.
Stop showing the cheapest supplier based on the cheapest room.
Properly select & show the cheapest room within each room range.
20% average gain by properly selecting the cheapest rooms sorted by type & specs.
A monitoring robot for booked rate survey until check-in.
Rates can fluctuate until check in.
For cancelable rooms, you periodically recheck until cancel
policy applies to get the last updated rate.
A mapped, qualified & deduplicated inventory
Get a detailed separation, curated data & description in 6 languages.
Have a improved UX with filters.
hotel mapping + room mapping.
To sort NET/BAR margins compared at the destination search level.
When mapping hundred hotels at the same time, you can sort GP margin by similar rooms grouped at destination level.
Case study: 3 bed-banks (NET) & 1 OTA (BAR), you search for a 2 bedrooms with a view at location
& sort by GP margin (NET/BAR) your options.
Right spot & push high profit opportunities
A personalized search engine made easy.
Click here to try an autocomplete example on our playground UX development page.
Offer the complete room range of the hotel, sorted out from your suppliers, deduplicated, with clean rewritten descriptions.
Add filters, personalization to your website. Create an auto-complete search engine working at hotel,
rental & room level, amenities, the view & so much more.
(we provide the auto-complete code for free)
how it works
-
all in 1 standardization tool
Natural Language Processing
NLP: Splitting a text
into sortable data
Our in house technology chunks a text into sections, scans terms, sorts out attributes.
Text Mining: It can be an hotel, rental or room name, a more complete text description.
A text size of a page can be passed.
(mapping time may vary)
Natural Language Understanding
NLU: Matching together
group similar
Rebuilding the room with cleansed data. Grouping similar rooms together.
Categorization: Matching is grouping the rooms based on normalized data using a defined number of attributes.
Our solution groups similar rooms together at multiple levels (customizable).
Natural Language Generation
NLG: Rewriting a description
Rewriting a cleansed, standardized description.
Highlighting: NLG is the counterpart of context analysis: its goal is to transform data into text.
Mimic Booking, Expedia on your website or set your own style.
Features & options
-
Covering a wide range of scenarios
6 languages IN
Our solution works simultaneously in 6 languages for the input, mixed or not. (EN ES PT FR IT DE)
Hoteliers, locals, often use their mother language in conjunction with English inside room description.
supplier | lang IN | room name |
---|---|---|
A | EN | 2 king suite sea view |
B | PT & ES | suite 2 camas king vista mar |
C | EN + FR | suite 2 king beds VUE MER |
6 languages OUT
Descriptions are translated into 6 languages.
mixed | EN + FR | suite 2 king beds VUE MER |
---|---|---|
1 | EN | Suite sea view, 2 king beds |
2 | ES | Suite vista mar, 2 camas king |
3 | PT | Suíte vista mar, 2 camas king |
3 | FR | Suite vue mer, 2 lits king |
3 | IT | Suite vista mare, 2 letti king |
3 | DE | Suite Meerblick, 2 Kingsize-Betten |
Custom description rewrite
Show & rewrite the room as you like it, customize your descriptions.
# | Rewrite it as needed |
---|---|
1 | Suite sea view, 2 kings. |
2 | Sea view suite, 2 kings. |
3 | Sea view, 2 king beds suite. |
Compatible with hotel mapping providers
Our solution is trans-vertical, agnostic & works seamlessly with a vast majority of actors.
This includes Giata, Vervotech, Gimmonix, bed-bank, OTA, Airbnb, GDS, B2B source or for channel & OTA on-boarding automation.
Covering all room's facets
More you map, less unknown.
Mapping a large number of attributes allows more "in context" analytic for high precision.
We are using 20 topics dictionaries splitted into 4 groups, depending on their importance for matching.
Customize your groups by keeping or removing an attribute, this is called granularity aggregation.
code | name | example |
---|---|---|
BDR | room generic | 1 bedroom, 2 bedrooms... |
TYP | room type | suite, loft, dormitory... |
CAT | class & category | standard, deluxe, corner... |
VEW | view, side, sight | sea/pool view, hill side, no view... |
BED | beds | double or twin, king bed, sofa... |
BLC | balcony, patio | balcony, terrace... |
CAP | capacity | single, double, triple... |
SGL | single use | 1 pax, single use... |
PAX | pax usage | no child, 2ad+1ch... |
SHR | anything shared | bathroom, bedroom... |
code | name | example |
---|---|---|
COK | cooking | kitchen... |
WWW | internet | wifi... |
CAR | parking | free parking... |
AMN | in room amenities | safe, minibar... |
FAC | out room facilities | spa, pool, lounge... |
SQF | surface | 30 m2/ 300 sqf.... |
code | name | example |
---|---|---|
MKG | marketing & name | stunning, art deco... |
CHA | Chains | marriott... |
FLR | floor | top floor, ground floor... |
ADA | Accessible | roll in... |
code | name | example |
---|---|---|
REF | refundable | no ref, refundable... |
CAN | cancelable | NC, cancelable... |
DEP | deposit | required... |
MBR | membership | members only... |
PLA | plan & promo | on sale, early birds... |
BRD | board | AI, half-board... |
SMK | smoking | no smoking, smoking... |
Similar rooms grouped
Granularity aggregation: It defines the number of attributes to use to match similarities.
Our solution returns an index per room, being its group. Customizable.
Case study 1: Match rooms according to type, class & view, stay similarly strict as original.
room code | room name | group index |
---|---|---|
120211 | superb suite dlx, lounge access,sea view & 2 king beds | SUITE DELUXE SEA VIEW |
ST.VX2.2KG | Suite superior vista mar, 2 king lounge | SUITE SUPERIOR SEA VIEW |
2DBL-STE-SEA | suite premier seavx 2 double. | SUITE PREMIER SEA VIEW |
Using NLU & parent categorization to match more. Optional & customizable.
Case study 2: Matching room range. In this example,
deluxe, superior & premier are regrouped in a common (SUP) group.
room code | room name | group index |
---|---|---|
143456 | superb suite dlx, lounge access,sea view & 2 king beds | SUITE (SUP) SEA VIEW |
ST.VX2.2KG | Suite superior vista mar, 2 fullbed lounge | SUITE (SUP) SEA VIEW |
2DBL-STE-SEA | premier suite sea vx 2 double bd. | SUITE (SUP) SEA VIEW |
Case study 3: Matching accessibility.
room code | room name | group index |
---|---|---|
3476654 | superb suite, lounge access,sea view & 2 king beds accessible | SUITE SEA VIEW ACCESSIBLE |
ST.VX.ADA | Suite vista mar, 2 king + lounge & roll in shower | SUITE SEA VIEW ACCESSIBLE |
2DBL-STE-SEA | suite sea vx 2 double bed ADA compliant. | SUITE SEA VIEW ACCESSIBLE |
Case study 4: Matching capacity.
room code | room name | group index |
---|---|---|
134564 | superb suite, lounge access,sea view for 2 adults | SUITE SEA VIEW (DOUBLE) |
ST.VX2.KG | Suite vista mar, king | SUITE SEA VIEW (DOUBLE) |
DBL-STE | double suite sea vx. | SUITE SEA VIEW (DOUBLE) |
Case study 5: Matching views.
room code | room name | group index |
---|---|---|
120211 | suite dlx sagrada familia view | SUITE DELUXE BASILICA VIEW |
ST.VX2.2KG | deluxe suite vista para la basilica | SUITE DELUXE BASILICA VIEW |
2DBL-STE-SEA | suite deluxe vue basilique | SUITE DELUXE BASILICA VIEW |
Vacation rental smart mapping
Vacation rentals are described like rooms and map the same way.
You clearly separate them from hotels & compare them with rooms.
We see 50% of listings being vacation rentals on high demand.
Hotel specs, style, marketing
Passing an hotel name or description, you map the type, chain, marketing & superlatives.
Categorize the hotel style, mood & map facilities & amenities.
Our solution allows to only select the appropriate attributes to be mapped depending on the application.
Envision global market margins
Compare 100 hotels at location.
Sort intermediation margin per room type & group at a large scale to refine the selection.
Right spot high margin offers & push them on your website.
Red flag / negation
Polarity is handled across attributes.
The API returns a red flag per topic if anything negative shows up. no view...
Anonymous funnel
Our service is source agnostic. name your supplier with a number or an alias.
According to the fields sent, the API automatically sort, match & group rooms accordingly.
Mapping %
2 important metrics to consider:
1 - Within each topic, what is the % coverage ?
It depends on the topic, some are easier than others.
For instance, views. You have the generic ones (beach, sea...) encountered in 90% of cases.
And you have the descriptive ones (broadway, monkey island...). we currently map 200k.
2 -How many topics are mapped ?
We map 20 topics. rooms having 100% curated content, receives a trust "footprint", useful in many applications.
We are reaching 80% of mapped rooms being 100% content mapped whatever is inside the description.
Trust level & leftover
The API returns a level of trust per room.
In the response, rooms being 100% curated are separated (80% of times).
Turnkey CSV cache
We provide the separation per column, the index groups, parent classification.
Use personalization and advanced filtering on your application.
Other business domains
Our technology is universal & works in many business domains.
Upload your own data & start mapping it.
Do you need to map camping ? tours or attractions ? Tell us more and we will set it up.
increased revenue, personalized search, boosted productivity
For any dynamic & static mapping jobs.
Our solution works seamlessly with your inventory & application.
Download a turnkey mapped database cache inventory.
What is NLP ? (wikipedia)
Fast integration
-
Any infra type & size
Our solution can be integrated in multiple ways
API
Simply connect to our API.
Direct
Dedicated ultra high speed cluster.
Cloud
In infra. Deploy within your AWS.
Innovation & development
Our in house NLP A.I. is a Natural Language Processing database-less SaaS.
It does not rely on traditional algorithms & is insensitive to the quantity of data to handle.
Any attributes dataset can be imported in it & mapped right away.
covering your business goals
We can tailor made you a 4000 rooms per sec handling 100 // calls per sec application.
A dedicated mapping/matching application customized to your needs.
Infrastructure & deployment
We know security, stability & secrecy is important for you and room mapping is at the heart of application.
Our solution is anonymous, autonomous, without dependencies or tiers.
powerful API ?
Fast integration & maximum automation. Instructions in 12 code languages. Result as CSV or JSON. Customer support, customization & training provided.
Current research
-
Inventory virtualization
on demand ID table mapping & maintenance
Traditionally, when you build an inventory, you are using a mapped ID table.
The benefits are obvious but by essence, an inventory is an additional layer to maintain.
Despite using modern technics like machine learning, ID table maintenance is power & time consuming.
All through, working, you end up using a few % of the maps between each update.
To drastically reduce inventory maintenance, mapping time & cost in general.
By being able to use virtualization mapping at search level, the goal is to refresh IDs & rooms on demand.
Either when creating a new inventory, or updating one.
You might think why would I need to constantly remap the same ? You do not, despite being possible.
You turn it on demand, use it in // of your search engine with your current inventory Ids to be updated.
Case study: you have 5000 best sellers, you need them "hot" with up to date content more frequently than 99% needing an update once a month.
To have a dynamically updated, maintenance-less & fresh inventory.
Our ID mapper is a new lightweight alternative for hospitality inventory creation.
To map ID + room codes simultaneously.
Mapping ID & rooms "on the fly".
No inventory maintenance needed.
Dynamic ID mapping at search level.
No ID or room codes cache needed.
ID, room codes & type mapping are done "live".
Data directly from suppliers' inventory.
No pre mapped ID service needed.
expedia, getaroom, gimmonix, goglobal, hotelbeds, innstant, mikitravel, ratehawk, sunhotels, tboholidays, airtours, booking.
Plus hotel & vacation rental separation at address.
Our Natural Language Processing A.I., adapted to ID mapping, eliminates false positive.
The goal are, first an error free mapping, second to reach +99%.
click below & see it for yourself.
Tell us what you would like to see on the interface & we will add it. Get free access.
info@room-matching.com
Real time API call demo
demo 2
-
cache usage for personalization
-
hotel + rental + room levels
Live rates compared & dynamic map zoom synchronized with database cache demo.
click below search box to try it live
Included features
Add-ons
Cost & usage
Included features
Cost & usage
Included features
Cost & usage
Included features
Cost & usage
You can use it for a variety of ML apps ( granular RM, geo-data market, rate predict, auto markup, TA %, ABS lvl meta search, OTA sentimental search... ).
The API response is a csv or Json dataset, turnkey for data scientists.
Train your custom model locally with PyTorch, on Google TensorFlow, AWS GPU instance.
please call.
Making data
work for you
Strategy design & tech roadmap
Cloud & infrastructure cost reduction
Third parties & tech mix requirements
Technical supervising & assistance
Artificial Intelligence & personalization
Advanced fintech design thinking
Inventory & content management
travel alliance consulting
Olivier Boinet
Hospitality big data expert
Travel Fintech NLP A.I. Pioneer
Glossary
Natural Language Processing
NLP:
Splitting a text
into sortable data
Our in house technology chunks a text into sections, scans terms, sorts out attributes.
Text Mining: It can be an hotel, rental or room name, a more complete text description.
A text size of a page can be passed.
(mapping time may vary)
Natural Language Understanding
NLU:
Matching together
group similar
Rebuilding the room with cleansed data. Grouping similar rooms together.
Categorization: Matching is grouping the rooms based on normalized data using a defined number of attributes.
Our solution groups similar rooms together at multiple levels (customizable).
Natural Language Generation
NLG:
Rewriting a description
Rewriting a cleansed, standardized description.
Highlighting: NLG is the counterpart of context analysis: its goal is to transform data into text.
Mimic Booking, Expedia on your website or set your own style.
Consulting services
Strategy
You can better weight your agreements with tiers. We set the tech roadmap & I help your team plan & synchronize integration.
Speeding production to market with an all in 1 platform has advantages but it will always cost you at the end.
Having an hybrid flexible application to gain more independence (and margin) & mitigate future costs is important.
Execution
You know the full story, pros & cons, strength & weakness of dependencies, mixed or not.
Each option weighted, checked for quality, inter-interoperability & cost vs benefit.
The customer search at location is his first contact with your hospitality data. You want quantity, quality & speed.
Production
We set up the right cloud hospitality management to be seamless with your application.
Are you an OTA, TA, TO, building a platform, a bed-bank ? Doing rate B.I., Machine Learning, heavy payload analytic ?
Each hospitality component has its own infra specificities to be cost effective. Mix them the right way to max out efficiency.
Personalization
I sort out & assess each component or dependency development potential of your project.
Development
Content aggregation & distribution, personalization @ search level, NLP search engine, databases classifications & structures.
A.I., NLP, NLU, Deep neural, LLM, OpenAI, ChatGPT, chatbot.
A.I. fintech
- automated ABS onboarding
- PyTorch tensorflow datasets
- Room range granular compSet
- Csv & json for data scientists
- Pre-trained & recursive models
- Fine tuned sales @ search level
- Room range rate prediction ML
About me
Having worked on mass market products, branding & behavioral marketing, my vision is fundamentally customer & cost oriented.
Also a computer sciences entrepreneur for decades, with extensive knowledge in cyber-security, core opcode dev, mastering multiple languages & environments.
Bringing smart intelligences to customer is at in the heart of work.
Applied to hospitality big data, I have developed room-matching.com
Working with me
By experience with inventions, I know innovation & success come from good ideas aggregation & shared knowledge & competences.
There is an old geek saying, "never trust a user keyboard input". When it comes to hospitality distribution & connectivity, call me Saint Thomas.
Current researches
Working on an automated inventory mapper, with a building online interface, to map at search level in real time, hotel ID & room codes.
I call it inventory virtualization. Available January 2024.
prototype:
https://meta.hotel-matching.com/IDmapping.html
travel dedicated LLMs
Machine learning has limitations.
Working on an hospitality local LLMs using a supervized NLP neural network.
Goal is to reach a high level of sentimental personalization across chatbot, speechtotext, search @ location, hotel & room levels.
We offer a new alternative mapping SaaS, working with any inventory. A more reactive & simpler way, independent from supplier changes or tiers. A new alternative for more independence.
- NLP, NLU, NLG a la carte.
- Customer & Personalization orientation.
- Mapping algorithms, code.
- Search engine, auto-complete.
- Database cache creation.
- Indexing, labeling, training.
- Evaluation, tips & tricks.
- Inventories aggregation.
travelalliance.com - room-matching.com - hotel-matching.com - 2024 - all right reserved