Upgrade to Pro — share decks privately, control downloads, hide ads and more …

不動産テックの見地からみたウェブサイエンスへの期待 / Expectation for the...

不動産テックの見地からみたウェブサイエンスへの期待 / Expectation for the Web Science from the Viewpoint of Real Estate Tech

JSAI SIG Web Science Open Seminar Vol.6 (May 26, 2018)
Webサイエンス研究会 オープンセミナー Vol.6

Kiyota Yoji, Ph.D.

May 26, 2018
Tweet

More Decks by Kiyota Yoji, Ph.D.

Other Decks in Technology

Transcript

  1. ෆಈ࢈ςοΫͷݟ஍͔ΒΈͨ ΢ΣϒαΠΤϯε΁ͷظ଴ L I F UL L L a bɹओ੮ݚڀһ

    ਗ਼ాɹཅ࢘ ΢ΣϒαΠΤϯεݚڀձ Φʔϓϯηϛφʔ Vol.6 2018.05.26 Copyright© LIFULL All Rights Reserved.
  2. ਗ਼ా ཅ࢘ גࣜձࣾLIFULLɹLIFULL Lab 1975೥෱Ԭݝੜ·Ε→େֶ(Ӄ)ˏژ౎ →ࢦಋڭһͷҟಈͰ౦ژ΁ ؔ৺෼໺: ࣗવݴޠॲཧԠ༻ → ݕࡧɾਪન

    → ৘ใϦςϥγʔ (ਤॻؗ) → ੜ׆ྖҬ (ෆಈ࢈ɺհޢ etc.) ܦྺ: େֶڭһ → ݉ۀͰىۀ → ͦͪΒ͕ຊۀʹ → ങऩ ओͳର֎త׆ಈ •  ਓ޻஌ೳֶձ ฤूҕһ •  ৘ใॲཧֶձ σʔλϕʔεγεςϜݚڀձ װࣄ •  WebDB Forum ࢈ֶ࿈ܞ୲౰װࣄ •  Code4Lib JAPANڞಉ୅දɹetc.
  3. LIFULL Lab (ϦοςϧϥϘϥτϦʔ) •  ϧʔπ͸2007೥ઃཱͷ ౦ژେֶൃελʔτ Ξοϓ (ג)Ϧοςϧ •  ෆಈ࢈ςοΫͳͲʹؔ

    ΘΔ࠷ઌ୺ٕज़ͷR&D •  Ϩίϝϯσʔγϣϯ •  AIؔ࿈ٕज़ʢσΟʔ ϓϥʔχϯάͳͲʣ •  ϢʔβʔΠϯλ ϑΣʔε (UX) •  VR / AR / MR গਓ਺ʢ2017೥10݄ݱ ࡏ 6໊ʣͰଟذʹΘͨ Δ෼໺ΛΧόʔ ৽ײ͓֮෦԰୳͠ΞϓϦ )0.&`4ϔϠαΫʂ ෦԰࡞ΓγϛϡϨʔγϣϯ (3*%73*$,
  4. ग़య: Manyika, James et al. Digital America: A tale of

    the haves and have-mores. McKinsey Global Institute. 2015 http://www.mckinsey.com/industries/high-tech/our-insights/digital-america-a-tale-of-the-haves-and-have-mores ෆಈ࢈ςοΫͷҐஔ͚ͮ ʢMcKinseyͷϨϙʔτʣ •  ଞͷʮXςοΫʯͱൺֱ ͢ΔͱσδλϧԽͷ౓߹ ͍͸தؒతͳҐஔ͚ͮ •  λʔήςΟϯά޿ࠂͳͲɺ ͢Ͱʹߴ౓ʹσʔλ׆༻ •  ҰํͰɺ઀٬΍಺ݟͳͲ ͸ਓʹཔΔ෦෼͕େ͖͍ •  ࠓޙͷσδλϧԽਁಁʹ ͱ΋ͳ͏มֵ͕ظ଴͞Ε Δ෼໺
  5. ෆಈ࢈ςοΫͷ՝୊ղܾʹ ޲͚ͯ •  ద੾ͳ෺݅Ձ஋ͷਪఆͷ೉͠͞ •  ໢ཏతͳ෺݅σʔλϕʔεͷະ੔උ ɹɹ→ ݚڀʹ࢖͑Δσʔλࢿݯ͕ඞཁ •  ফඅऀχʔζ೺Ѳͷ೉͠͞

    ɹɹ→ ڠௐϑΟϧλϦϯάͳͲͷ൚༻తͳਪન ɹɹɹΞϧΰϦζϜͱ͸ҟͳΔϞσϧͷݚڀ͕ඞཁ •  Virtual Reality (VR) ׆༻ͷน ɹɹ→ ͢Ͱʹଘࡏ͢Δσʔλʢࣸਅ΍ؒऔΓਤʣ ɹɹɹ͔Β3DσʔλΛੜ੒Ͱ͖ͳ͍͔ʁ
  6. ࠃ಺ͷ௞ିෆಈ࢈෺݅σʔλ ໿530ສ݅ •  ॴࡏ஍ (༣ศ൪߸ɺ࠷دΓӺͳͲ) •  ௞ྉɺ໘ੵɺங೥਺ɺ෦԰λΠϓ etc. •  ݐ෺ߏ଄

    (໦଄ɺమࠎɺమےίϯΫϦʔτ etc.) •  ֤छͩ͜ΘΓ৚݅ (ϖοτՄɺָثɺΧ΢ ϯλʔΩονϯɺόεɾτΠϨผ etc.) 物件画像 約8300万点 間取り図 約510万点 重厚な感じの エントランス 日当たりの 良いリビング •  国立情報学研究所(NII) の協力を得て2015年11 月より提供開始 •  学術機関の研究者であ れば無料で利用可能 LIFULL HOME’Sσʔληοτ
  7. JSAI 2018 OS-17ʮෆಈ࢈ͱAIʯ ʢ6/6 15:00-17:20ʣ 1. Ϣʔβɾ෺݅ಛ௃ͱؒऔΓಛ௃Λ༻͍ͨᅂ޷༧ଌ ʓՃ౻ ঘًɺࢁ࡚ ढ़඙ɺ૬ᖒ

    ਗ਼੖ʢ౦ژେֶʣ େ඿ ؽ ඒʢגࣜձࣾiettyʣ 2. ෆಈ࢈ϙʔλϧαΠτͰͷϢʔβͷજࡏతᅂ޷ͷநग़ ʓ஑্ ݦਅʢגࣜձࣾϦΫϧʔτॅ·͍Χϯύχʔʣ ҏ ౻ തయʢژ౎େֶʣ ໺ଜ ᚸฏʢגࣜձࣾϦΫϧʔτॅ· ͍Χϯύχʔʣ 3. Webෆಈ࢈σʔλΛ༻͍ۭͨ෺͕݅ຒ·ΔભҠʹؔ͢Δଟ มྔղੳ ʓ౉ᬑ ൏࢙ʢ౷ܭ਺ཧݚڀॴʣ Ұ౻ ༟ʢ௕࡚େֶʣ ླ໦ խਓʢUDΞηοτόϦϡΤʔγϣϯʣ ࢁԼ ஐࢤʢ౷ܭ਺ ཧݚڀॴʣ 4. Bottleneckಛ௃ྔΛ༻͍ͨෆಈ࢈ը૾ͷ෼ྨ ʓ๛ݪ ༏ʢ౦ژཧՊେֶʣ ໳ ༸ҰʢΞοτϗʔϜגࣜձ ࣾʣ ࢁ࡚ ढ़඙ʢ౦ژେֶʣ ౻৿ ਐɺଠݪ ҭ෉ʢ౦ژཧ Պେֶʣ
  8. ICMR 2018ซઃϫʔΫγϣοϓ Multimedia for RETech (6/11 ޕޙ) Users’ Preference Prediction

    of Real Estates Featuring Floor Plan Analysis using FloorNet (Naoki Kato, Toshihiko Yamasaki, Kiyoharu Aizawa and Takemi Ohama) Visual Estimation of Building Condition with Patch-level ConvNets (David Koch, Miroslav Despotovic, Muntaha Sakeena, Mario Doeller and Matthias Zeppelzauer) Proposition of VR-MR Hybrid System for Sharing Living-in-Room (Naoki Murata, Satoshi Suga, Eichi Takaya, Satoshi Ueno, Yoji Kiyota and Satoshi Kurihara) Apartment Structure Estimation Using Fully Convolutional Networks and Graph Model (Toshihiko Yamasaki, Jin Zhang and Yuki Takada) Fast Semi-dense Depth Map Estimation (Ilya Makarov, Alisa Korinevskaya and Vladimir Aliev)
  9. ౎ࢢ͕ଟ༷ੑΛ΋ͭͨΊͷ4৚݅ 1.  ࠞ߹༻్ͷඞཁੑ: ஍͕۠ɺ࠷௿2ͭҎ্ͷػೳ ΛՌͨ͢͜ͱ 2.  খ͞ͳ֗۠: ֗۠͸୹͘ɺ֗࿏΍֯Λۂ͕Δػձ ͕ଟ͍͜ͱ 3. 

    ݹ͍ݐ෺ͷඞཁੑ: ݹ͞΍৚͕݅ҟͳΔ֤छͷݐ ෺͕ࠞࡏ͢Δ͜ͱ 4.  ີूͷඞཁੑ: े෼ͳີ౓Ͱਓ͕͍Δ͜ͱ