ã°ã©ãã¢ããªãã£ã¯ã¹ãšãããã¯ãŒã¯åæã®åãæ¢æ±ããæ¥ç¶ãããããŒã¿ã«é ããããã¿ãŒã³ãé¢ä¿æ§ãã€ã³ãµã€ããæããã«ããŸããå®çšçãªå¿çšãã¢ã«ãŽãªãºã ãå®äžçã®äºäŸãåŠã³ãŸãããã
ã°ã©ãã¢ããªãã£ã¯ã¹ïŒãããã¯ãŒã¯åæã§ã€ã³ãµã€ããè§£ãæãã
仿¥ã®çžäºæ¥ç¶ãããäžçã§ã¯ãããŒã¿ã¯ãŸããŸãé¢ä¿æ§ã®åœ¢ã§ååšããŠããŸãããœãŒã·ã£ã«ãããã¯ãŒã¯ãããµãã©ã€ãã§ãŒã³ãŸã§ããããã®ã€ãªãããçè§£ããããšã¯ãç«¶äºäžã®åªäœæ§ãç²åŸããè€éãªåé¡ã解決ããæ å ±ã«åºã¥ããæææ±ºå®ãè¡ãäžã§æ¥µããŠéèŠã§ããããã§æŽ»èºããã®ãããããã¯ãŒã¯åæãåååãšããã°ã©ãã¢ããªãã£ã¯ã¹ã§ãããã®èšäºã§ã¯ãã°ã©ãã¢ããªãã£ã¯ã¹ã®æŠå¿µãå¿çšãã¢ã«ãŽãªãºã ããããŠæ§ã ãªæ¥çã«ãããå®äžçã®äºäŸãæ¢æ±ããå æ¬çãªæŠèŠãæäŸããŸãã
ã°ã©ãã¢ããªãã£ã¯ã¹ãšã¯ïŒ
ã°ã©ãã¢ããªãã£ã¯ã¹ãšã¯ãããŒãïŒãšã³ãã£ãã£ïŒãšãšããžïŒé¢ä¿æ§ïŒã§æ§æãããã°ã©ããšããŠè¡šçŸãããããŒã¿ãåæããããã»ã¹ã§ããããŒãã«å ã®æ§é åããŒã¿ã«çŠç¹ãåœãŠãåŸæ¥ã®ãªã¬ãŒã·ã§ãã«ããŒã¿ããŒã¹ãšã¯ç°ãªããã°ã©ãããŒã¿ããŒã¹ãšã¢ããªãã£ã¯ã¹ã¯ããŒã¿ãã€ã³ãéã®æ¥ç¶ãéèŠããŸãããããã¯ãŒã¯åæã¯ããããã®ã°ã©ãæ§é ãåæããããã«äœ¿çšãããäžé£ã®æè¡ã§ãã
ã°ã©ãã¢ããªãã£ã¯ã¹ã®äž»èŠãªæŠå¿µã«ã¯ã以äžã®ãã®ããããŸãïŒ
- ããŒãïŒäººã補åãçµç¹ãå Žæãªã©ã®ãšã³ãã£ãã£ã衚ããŸãã
- ãšããžïŒããŒãéã®é¢ä¿æ§ã衚ããŸããåæ ãè³Œå ¥ãã³ãã¥ãã±ãŒã·ã§ã³ãªã©ãããã«ããããŸãããšããžã¯æåïŒäžæ¹åïŒãŸãã¯ç¡åïŒåæ¹åïŒã§ãããããããã£ãéã¿ãæã€ããšãã§ããŸãã
- ã°ã©ãïŒããŒããšãšããžã®éåã§ãã
- ã°ã©ãããŒã¿ããŒã¹ïŒã°ã©ãããŒã¿ãå¹ççã«ä¿åããã¯ãšãªãå®è¡ããããã«èšèšãããç¹æ®ãªããŒã¿ããŒã¹ã§ããäŸãšããŠãNeo4jãAmazon NeptuneãJanusGraphãªã©ããããŸãã
ã°ã©ãã¢ããªãã£ã¯ã¹ã«ãã£ãŠã次ã®ããšãå¯èœã«ãªããŸãïŒ
- ãã¿ãŒã³ãšé¢ä¿æ§ã®ç¹å®ïŒããŒã¿å ã«é ãããã€ãªãããäŸåé¢ä¿ãçºèŠããŸãã
- ãããã¯ãŒã¯æ§é ã®çè§£ïŒãããã¯ãŒã¯å šäœã®çµç¹ãæ å ±ã®æµããåæããŸãã
- å°æ¥ã®è¡åã®äºæž¬ïŒãããã¯ãŒã¯ã®ç¹æ§ãçšããŠãã¬ã³ããçµæãäºæž¬ããŸãã
- æææ±ºå®ã®æ¹åïŒæŠç¥çèšç»ãæ¥åå¹çã«åœ¹ç«ã€ã€ã³ãµã€ããåŸãŸãã
ã°ã©ãã¢ããªãã£ã¯ã¹ãéèŠã§ããçç±
ã°ã©ãã¢ããªãã£ã¯ã¹ã®åã¯ãåŸæ¥ã®ããŒã¿åæææ³ã§ã¯èŠéããããã¡ãªã€ã³ãµã€ããæããã«ã§ããç¹ã«ãããŸãããªããã®éèŠæ§ãå¢ããŠããã®ããçç±ã¯ä»¥äžã®éãã§ãïŒ
- æ¥ç¶ãããããŒã¿ã¯è³ãæã«ååšããïŒãœãŒã·ã£ã«ã¡ãã£ã¢ãããã¯ãŒã¯ããéèååŒãŸã§ã仿¥çæãããããŒã¿ã®å€ãã¯æ¬è³ªçã«æ¥ç¶ãããŠããŸããã°ã©ãã¢ããªãã£ã¯ã¹ã¯ããã®æ¥ç¶ãããããŒã¿ã广çã«åæããããã®ããŒã«ãæäŸããŸãã
- é ããé¢ä¿æ§ã®çºèŠïŒã°ã©ãåæã¯ãåŸæ¥ã®ãªã¬ãŒã·ã§ãã«ããŒã¿ããŒã¹ã®ã¯ãšãªãçµ±èšåæã§ã¯æããã«ãªããªããããªãèªæã§ãªãé¢ä¿æ§ãèŠã€ããããšã«é·ããŠããŸããããã«ããã顧客è¡åã®çè§£ãäžæ£æ€åºãç§åŠççºèŠã«ããããã¬ãŒã¯ã¹ã«ãŒãããããããå¯èœæ§ããããŸãã
- äºæž¬ã¢ããªã³ã°ã®åŒ·åïŒãããã¯ãŒã¯æ å ±ãäºæž¬ã¢ãã«ã«çµã¿èŸŒãããšã§ããã®ç²ŸåºŠãšæå¹æ§ãåäžãããããšãã§ããŸããäŸãã°ã顧客ã®ç€ŸäŒçã€ãªãããç¥ãããšã§ãè§£çŽäºæž¬ãæ¹åã§ããŸãã
- æææ±ºå®æ¯æŽã®åäžïŒã°ã©ã衚çŸã®èŠèŠçã§çŽæçãªæ§è³ªã«ãããè€éãªé¢ä¿æ§ãçè§£ããã¹ããŒã¯ãã«ããŒã«ã€ã³ãµã€ããäŒãããããªããŸãã
äž»èŠãªã°ã©ãã¢ããªãã£ã¯ã¹æè¡ãšã¢ã«ãŽãªãºã
ã°ã©ãã¢ããªãã£ã¯ã¹ã§ã¯ããããã¯ãŒã¯ããŒã¿ããææçŸ©ãªã€ã³ãµã€ããæœåºããããã«ãæ§ã ãªæè¡ãšã¢ã«ãŽãªãºã ãçšããããŸããæãéèŠãªãã®ã«ã¯ä»¥äžã®ãããªãã®ããããŸãïŒ
äžå¿æ§ææš
äžå¿æ§ææšã¯ããããã¯ãŒã¯å ã§ã®äœçœ®ãšæ¥ç¶ã«åºã¥ããŠãæãéèŠãªããŒããç¹å®ããŸããäžè¬çãªäžå¿æ§ææšã«ã¯ä»¥äžãå«ãŸããŸãïŒ
- 次æ°äžå¿æ§ïŒDegree CentralityïŒïŒããŒããæã€çŽæ¥çãªæ¥ç¶ã®æ°ã枬å®ããŸããæ¬¡æ°äžå¿æ§ãé«ãããŒãã¯ããã®çŽæ¥çãªè¿åã§éåžžã«å€ãã®æ¥ç¶ãæã¡ã圱é¿åã倧ããã§ãã
- åªä»äžå¿æ§ïŒBetweenness CentralityïŒïŒããããŒããä»ã®2ã€ã®ããŒãéã®æççµè·¯äžã«äœçœ®ããåæ°ãæž¬å®ããŸããåªä»äžå¿æ§ãé«ãããŒãã¯ããããã¯ãŒã¯å ã§ããªããžãã²ãŒãããŒããŒãšããŠæ©èœããŸãã
- è¿æ¥äžå¿æ§ïŒCloseness CentralityïŒïŒããããŒããããããã¯ãŒã¯å ã®ä»ã®ãã¹ãŠã®ããŒããŸã§ã®å¹³åè·é¢ã枬å®ããŸããè¿æ¥äžå¿æ§ãé«ãããŒãã¯ããããã¯ãŒã¯ã®ãã¹ãŠã®éšåãã容æã«ã¢ã¯ã»ã¹ã§ããŸãã
- åºæãã¯ãã«äžå¿æ§ïŒEigenvector CentralityïŒïŒããŒãã®åœ±é¿åãããã®é£äººã®åœ±é¿åã«åºã¥ããŠæž¬å®ããŸããéèŠãªããŒãã«æ¥ç¶ãããŠããããŒãã¯ãéèŠã§ãããšèŠãªãããŸããGoogleã䜿çšããPageRankã¯ãåºæãã¯ãã«äžå¿æ§ã®äžçš®ã§ãã
äŸïŒãœãŒã·ã£ã«ãããã¯ãŒã¯ã«ãããŠã次æ°äžå¿æ§ãé«ã人ã¯äººæ°è ãšèŠãªããããããããŸããããåªä»äžå¿æ§ãé«ã人ã¯éèŠãªæ å ±ã³ãã¯ã¿ãŒã仲ä»è ã§ããå¯èœæ§ããããŸãã
ã³ãã¥ããã£æ€åº
ã³ãã¥ããã£æ€åºã¢ã«ãŽãªãºã ã¯ããããã¯ãŒã¯ã®ä»ã®éšåãããäºãã«å¯ã«æ¥ç¶ãããŠããããŒãã®ã°ã«ãŒããç¹å®ããŸãããããã®ã°ã«ãŒãã¯ãé¢é£ãããšã³ãã£ãã£ã®ã³ãã¥ããã£ãã¯ã©ã¹ã¿ãŒã衚ããŸãã
äžè¬çãªã³ãã¥ããã£æ€åºã¢ã«ãŽãªãºã ã«ã¯ä»¥äžãå«ãŸããŸãïŒ
- ã«ãŒãã³æ³ïŒLouvain AlgorithmïŒïŒãããã¯ãŒã¯ã®ã¢ãžã¥ã©ãªãã£ïŒã³ãã¥ããã£å ã®æ¥ç¶å¯åºŠãšã³ãã¥ããã£éã®æ¥ç¶å¯åºŠã®æ¯èŒïŒãå埩çã«æé©åãã貪欲æ³ã¢ã«ãŽãªãºã ã§ãã
- ã©ãã«äŒææ³ïŒLabel Propagation AlgorithmïŒïŒåããŒãã¯æåã«äžæã®ã©ãã«ãå²ãåœãŠããããã®åŸãé£äººã®äžã§æãé »ç¹ãªã©ãã«ã«äžèŽããããã«èªèº«ã®ã©ãã«ãå埩çã«æŽæ°ããŸããåãã©ãã«ãæã€ããŒããã¯ã©ã¹ã¿ãŒåããããšã§ã³ãã¥ããã£ã圢æãããŸãã
- ã¬ãŒãã³ã»ãã¥ãŒãã³æ³ïŒGirvan-Newman AlgorithmïŒïŒåªä»äžå¿æ§ãæãé«ããšããžãå埩çã«åé€ãããããã¯ãŒã¯ãåŸã ã«å°ããåå²ããŠããåå²åã®ã¢ã«ãŽãªãºã ã§ãã
äŸïŒé¡§å®¢ãããã¯ãŒã¯ã«ãããŠãã³ãã¥ããã£æ€åºã¯åæ§ã®è³Œè²·ç¿æ £ãèå³ãæã€é¡§å®¢ã°ã«ãŒããç¹å®ããã¿ãŒã²ãããçµã£ãããŒã±ãã£ã³ã°ãã£ã³ããŒã³ãå¯èœã«ããŸãã
çµè·¯æ¢çŽ¢ã¢ã«ãŽãªãºã
çµè·¯æ¢çŽ¢ã¢ã«ãŽãªãºã ã¯ããããã¯ãŒã¯å ã®2ã€ã®ããŒãéã®æçãŸãã¯æãå¹ççãªçµè·¯ãèŠã€ããŸãããããã®ã¢ã«ãŽãªãºã ã¯ãã«ãŒãã£ã³ã°ãæšå¥šããããã¯ãŒã¯æé©åã«åœ¹ç«ã¡ãŸãã
äžè¬çãªçµè·¯æ¢çŽ¢ã¢ã«ãŽãªãºã ã«ã¯ä»¥äžãå«ãŸããŸãïŒ
- ãã€ã¯ã¹ãã©æ³ïŒDijkstra's AlgorithmïŒïŒãšããžã«é¢é£ããã³ã¹ããè·é¢ãããéã¿ä»ãã°ã©ãã§ã2ã€ã®ããŒãéã®æççµè·¯ãèŠã€ããŸãã
- A*ïŒãšãŒã¹ã¿ãŒïŒæ¢çŽ¢ã¢ã«ãŽãªãºã ïŒãã€ã¯ã¹ãã©æ³ãæ¡åŒµãããã®ã§ããã¥ãŒãªã¹ãã£ã¯ã¹ãçšããŠæ¢çŽ¢ãèªå°ããå€§èŠæš¡ãªã°ã©ãã«å¯ŸããŠããå¹ççã«åäœããŸãã
- æççµè·¯ã¢ã«ãŽãªãºã ïŒéã¿ãªãã°ã©ãïŒïŒå¹ åªå æ¢çŽ¢ïŒBFSïŒã®ãããªã¢ã«ãŽãªãºã ã¯ããã¹ãŠã®ãšããžãåãéã¿ãæã€ã°ã©ãã§æççµè·¯ãå¹ççã«èŠã€ããããšãã§ããŸãã
äŸïŒç©æµãããã¯ãŒã¯ã«ãããŠãçµè·¯æ¢çŽ¢ã¢ã«ãŽãªãºã ã¯ååã®é éã«æé©ãªã«ãŒããæ±ºå®ããç§»åæéãšã³ã¹ããæå°åããããšãã§ããŸãã
ãªã³ã¯äºæž¬
ãªã³ã¯äºæž¬ã¢ã«ãŽãªãºã ã¯ãæ¢åã®ãããã¯ãŒã¯æ§é ã«åºã¥ããŠã2ã€ã®ããŒãéã«å°æ¥çãªæ¥ç¶ãçãŸããå¯èœæ§ãäºæž¬ããŸããããã¯ãæšå¥šã·ã¹ãã ããœãŒã·ã£ã«ãããã¯ãŒã¯åæãäžæ£æ€åºã«åœ¹ç«ã¡ãŸãã
äžè¬çãªãªã³ã¯äºæž¬æè¡ã«ã¯ä»¥äžãå«ãŸããŸãïŒ
- å ±éã®é£äººïŒCommon NeighborsïŒïŒ2ã€ã®ããŒããå ±éããŠæã€é£äººãå€ãã»ã©ãããããæ¥ç¶ã圢æããå¯èœæ§ãé«ããªããŸãã
- ãžã£ãã«ãŒãä¿æ°ïŒJaccard IndexïŒïŒ2ã€ã®ããŒãã®é£äººã»ããéã®é¡äŒŒåºŠã枬å®ããŸãã
- åªå çéžæïŒPreferential AttachmentïŒïŒããå€ãã®æ¥ç¶ãæã€ããŒãã¯ãæ°ããæ¥ç¶ãåŒãä»ãããããšããèãæ¹ã§ãã
äŸïŒãœãŒã·ã£ã«ãããã¯ãŒã¯ã«ãããŠããªã³ã¯äºæž¬ã¯çžäºã®ã€ãªãããå ±éã®èå³ã«åºã¥ããŠæ°ããåäººãææ¡ããããšãã§ããŸãã
ã°ã©ãé¡äŒŒåºŠ
ã°ã©ãé¡äŒŒåºŠã¢ã«ãŽãªãºã ã¯ã2ã€ã®ã°ã©ããŸãã¯ãµãã°ã©ãéã®æ§é çãªé¡äŒŒæ§ã枬å®ããŸããããã¯ãé¡äŒŒãããã¿ãŒã³ã®ç¹å®ããããã¯ãŒã¯ã®æ¯èŒãã°ã©ãã®ã¯ã©ã¹ã¿ãªã³ã°ã«åœ¹ç«ã¡ãŸãã
äžè¬çãªã°ã©ãé¡äŒŒåºŠå°ºåºŠã«ã¯ä»¥äžãå«ãŸããŸãïŒ
- ã°ã©ãç·šéè·é¢ïŒGraph Edit DistanceïŒïŒäžæ¹ã®ã°ã©ããããäžæ¹ã®ã°ã©ãã«å€æããããã«å¿ èŠãªç·šéæäœïŒããŒããŸãã¯ãšããžã®æ¿å ¥/åé€ïŒã®æå°åæ°ã
- æå€§å ±ééšåã°ã©ãïŒMaximum Common SubgraphïŒïŒäž¡æ¹ã®ã°ã©ãã«ååšããæå€§ã®ãµãã°ã©ãã
- ã°ã©ãã«ãŒãã«ïŒGraph KernelsïŒïŒã«ãŒãã«é¢æ°ã䜿çšããŠãã°ã©ãã®æ§é çç¹åŸŽã«åºã¥ããŠã°ã©ãéã®é¡äŒŒæ§ã枬å®ããŸãã
äŸïŒãã€ãªã€ã³ãã©ããã£ã¯ã¹ã«ãããŠãã°ã©ãé¡äŒŒåºŠã¯ã¿ã³ãã¯è³ªçžäºäœçšãããã¯ãŒã¯ãæ¯èŒããé¡äŒŒããæ©èœãæã€ã¿ã³ãã¯è³ªãç¹å®ããããã«äœ¿çšã§ããŸãã
ã°ã©ãã¢ããªãã£ã¯ã¹ã®å¿çš
ã°ã©ãã¢ããªãã£ã¯ã¹ã¯ãå¹ åºãæ¥çãåéã§å¿çšãããŠããŸãã以äžã«ããã€ãã®æ³šç®ãã¹ãäŸãæããŸãïŒ
ãœãŒã·ã£ã«ãããã¯ãŒã¯åæ
ãœãŒã·ã£ã«ãããã¯ãŒã¯åæïŒSNAïŒã¯ãã°ã©ãã¢ããªãã£ã¯ã¹ã®æãããç¥ãããå¿çšã®äžã€ã§ãã人ãçµç¹ããã®ä»ã®ãšã³ãã£ãã£ã®ãããã¯ãŒã¯å ã§ã®ç€ŸäŒçé¢ä¿ãçžäºäœçšãåæããŸãã
äŸïŒ
- ã€ã³ãã«ãšã³ãµãŒã®ç¹å®ïŒäžå¿æ§ææšã«åºã¥ããŠããœãŒã·ã£ã«ãããã¯ãŒã¯ã§æã圱é¿åã®ãã人ç©ãç¹å®ããŸããããã¯ãã¿ãŒã²ããããŒã±ãã£ã³ã°ãå ¬è¡è¡çãã£ã³ããŒã³ã«å©çšã§ããŸãã
- ã³ãã¥ããã£æ€åºïŒå ±éã®èå³ãæå±ãæã€äººã ã®ã°ã«ãŒããç¹å®ããŸããããã¯ãã¿ãŒã²ããåºåãç€ŸäŒæŽ»åã«å©çšã§ããŸãã
- ãœãŒã·ã£ã«ãããã¯ãŒã¯ããŒã±ãã£ã³ã°ïŒæ å ±ããœãŒã·ã£ã«ãããã¯ãŒã¯ãéããŠã©ã®ããã«åºããããçè§£ããããã«å¿ããŠããŒã±ãã£ã³ã°æŠç¥ãæé©åããŸãã
äžæ£æ€åº
ã°ã©ãã¢ããªãã£ã¯ã¹ã¯ãéèååŒãä¿éºéè«æ±ããã®ä»ã®ããŒã¿ã«ãããç°åžžãªãã¿ãŒã³ãé¢ä¿æ§ãç¹å®ããããšã«ãããäžæ£è¡çºã®æ€åºã«éåžžã«å¹æçã§ãã
äŸïŒ
- äžæ£ãªã³ã°ã®ç¹å®ïŒäžæ£ãå ±è¬ããŠããå人ãçµç¹ã®ã°ã«ãŒããæ€åºããŸãã
- ç°åžžæ€åºïŒéåžžãšã¯ç°ãªãååŒã掻åãç¹å®ããäžæ£è¡çºã®å¯èœæ§ã瀺åããŸãã
- ãªã³ã¯åæïŒçãããè©æ¬ºåž«éã®ã€ãªããã远跡ããé ããé¢ä¿æ§ãæããã«ããäžæ£ãããã¯ãŒã¯å šäœãæçºããŸãã
æšå¥šã·ã¹ãã
ã°ã©ãã¢ããªãã£ã¯ã¹ã¯ããŠãŒã¶ãŒãã¢ã€ãã ããã®ä»ã®ãšã³ãã£ãã£éã®é¢ä¿æ§ã掻çšããŠãããŒãœãã©ã€ãºãããæšå¥šãæäŸããããšã§ãæšå¥šã·ã¹ãã ã匷åã§ããŸãã
äŸïŒ
- è£œåæšå¥šïŒãŠãŒã¶ãŒã®éå»ã®è³Œå ¥å±¥æŽãé²èЧ履æŽã瀟äŒçã€ãªããã«åºã¥ããŠè£œåãæšå¥šããŸãã
- æ ç»æšå¥šïŒãŠãŒã¶ãŒã®è©äŸ¡ãã¬ãã¥ãŒãé¡äŒŒãŠãŒã¶ãŒã®å¥œã¿ã«åºã¥ããŠæ ç»ãæšå¥šããŸãã
- å人æšå¥šïŒçžäºã®ã€ãªãããå ±éã®èå³ã«åºã¥ããŠæ°ããåäººãææ¡ããŸãã
ãµãã©ã€ãã§ãŒã³æé©å
ã°ã©ãã¢ããªãã£ã¯ã¹ã¯ããµãã©ã€ãã§ãŒã³ãã¢ãã«åãæé©åããããã«äœ¿çšã§ããå¹çæ§ã®åäžãã³ã¹ãã®åæžããªã¹ã¯ã®è»œæžã«è²¢ç®ããŸãã
äŸïŒ
- ããã«ããã¯ã®ç¹å®ïŒé å»¶ãäžæãçºçãããããµãã©ã€ãã§ãŒã³ã®éèŠãªãã€ã³ããç¹å®ããŸãã
- ã«ãŒãæé©åïŒååã®èŒžéã«æé©ãªã«ãŒããæ±ºå®ããç§»åæéãšã³ã¹ããæå°åããŸãã
- ãªã¹ã¯ç®¡çïŒãµãã©ã€ãã§ãŒã³ã®æœåšçãªè匱æ§ãç¹å®ããç·©åæŠç¥ãçå®ããŸãã
ãã¬ããžã°ã©ã
ãã¬ããžã°ã©ãã¯ãç¥èãã°ã©ãããŒã¹ã§è¡šçŸãããã®ã§ã質åå¿çãæ å ±æ€çŽ¢ãã»ãã³ãã£ãã¯æ€çŽ¢ãªã©ãæ§ã ãªå¿çšã«å©çšã§ããŸããGoogleãFacebookã®ãããªäŒæ¥ã¯ããã¬ããžã°ã©ããåºç¯å²ã«æŽ»çšããŠããŸãã
äŸïŒ
- ã»ãã³ãã£ãã¯æ€çŽ¢ïŒæ€çŽ¢èªéã®æå³ãšé¢ä¿æ§ãçè§£ããããé¢é£æ§ã®é«ãæ€çŽ¢çµæãæäŸããŸãã
- 質åå¿çïŒãã¬ããžã°ã©ãäžã§æšè«ãè¡ãããšã§ãè€éãªè³ªåã«çããŸãã
- ããŒã¿çµ±åïŒè€æ°ã®ãœãŒã¹ããã®ããŒã¿ãçµ±äžããããã¬ããžã°ã©ãã«çµ±åããŸãã
ãã«ã¹ã±ã¢
ã°ã©ãã¢ããªãã£ã¯ã¹ã¯ãåµè¬ããæ£è ã±ã¢ãŸã§ããã«ã¹ã±ã¢ã«ãããŠãŸããŸãéèŠãªåœ¹å²ãæãããŠããŸãã
äŸïŒ
- åµè¬ïŒã¿ã³ãã¯è³ªçžäºäœçšãããã¯ãŒã¯ãçŸæ£çµè·¯ãåæããããšã§ãæœåšçãªåµè¬ã¿ãŒã²ãããç¹å®ããŸãã
- åå¥åå»çïŒæ£è å人ã®éºäŒçæ§æãå»çå±¥æŽããœãŒã·ã£ã«ãããã¯ãŒã¯ã«åºã¥ããŠæ²»çèšç»ã調æŽããŸãã
- ææççºçã®æ€åºïŒãœãŒã·ã£ã«ãããã¯ãŒã¯ãç§»åãã¿ãŒã³ãåæããããšã§ãææçã®æ¡å€§ã远跡ããŸãã
ã°ã©ãã¢ããªãã£ã¯ã¹ã®ããã®ããŒã«ãšãã¯ãããžãŒ
ã°ã©ãã¢ããªãã£ã¯ã¹ãå®è¡ããããã«ã¯ãå°éã®ã°ã©ãããŒã¿ããŒã¹ããæ±çšã®ããŒã¿ãµã€ãšã³ã¹ãã©ãããã©ãŒã ãŸã§ãããã€ãã®ããŒã«ãšãã¯ãããžãŒãå©çšå¯èœã§ãã
ã°ã©ãããŒã¿ããŒã¹
ã°ã©ãããŒã¿ããŒã¹ã¯ãã°ã©ãããŒã¿ãå¹ççã«ä¿åããã¯ãšãªãå®è¡ããããã«ç¹å¥ã«èšèšãããŠããŸããã°ã©ãæ§é ãšã¢ã«ãŽãªãºã ããã€ãã£ãã«ãµããŒãããŠãããããã°ã©ãã¢ããªãã£ã¯ã¹ã®å¿çšã«æé©ã§ãã
人æ°ã®ã°ã©ãããŒã¿ããŒã¹ã«ã¯ä»¥äžãå«ãŸããŸãïŒ
- Neo4jïŒè±å¯ãªæ©èœã»ãããšåŒ·åãªã³ãã¥ããã£ãæã€ãäž»èŠãªã°ã©ãããŒã¿ããŒã¹ã
- Amazon NeptuneïŒAmazon Web ServicesãæäŸããããã«ãããŒãžãã®ã°ã©ãããŒã¿ããŒã¹ãµãŒãã¹ã
- JanusGraphïŒè€æ°ã®ã¹ãã¬ãŒãžããã¯ãšã³ãããµããŒãããã忣åã®ãªãŒãã³ãœãŒã¹ã°ã©ãããŒã¿ããŒã¹ã
- Microsoft Azure Cosmos DBïŒã°ã©ãããŒã¿ããµããŒããããã°ããŒãã«ã«åæ£ããããã«ãã¢ãã«ããŒã¿ããŒã¹ãµãŒãã¹ã
ã°ã©ãã¢ããªãã£ã¯ã¹ãã©ãããã©ãŒã
ã°ã©ãã¢ããªãã£ã¯ã¹ãã©ãããã©ãŒã ã¯ãã°ã©ãããŒã¿ã®ç®¡çãåæãå¯èŠåã®ããã®å æ¬çãªããŒã«ãšæ©èœãæäŸããŸãã
äŸïŒ
- TigerGraphïŒè¶ 䞊åã®ã°ã©ãããŒã¿ããŒã¹ããã³ã¢ããªãã£ã¯ã¹ãã©ãããã©ãŒã ã
- GraphistryïŒã°ã©ãããŒã¿ã®ããã®ããžã¥ã¢ã«èª¿æ»ãã©ãããã©ãŒã ã
- GephiïŒãªãŒãã³ãœãŒã¹ã®ã°ã©ãå¯èŠåã»åæãœãããŠã§ã¢ã
ããã°ã©ãã³ã°èšèªãšã©ã€ãã©ãª
å€ãã®ããã°ã©ãã³ã°èšèªãšã©ã€ãã©ãªããã°ã©ãã¢ããªãã£ã¯ã¹ããµããŒãããŠããŸãã
äŸïŒ
- PythonïŒäººæ°ã®ã©ã€ãã©ãªã«ã¯NetworkX, igraph, Graph-toolããããŸãã
- RïŒigraphããã±ãŒãžã¯å æ¬çãªã°ã©ãåææ©èœãæäŸããŸãã
- JavaïŒApache TinkerPopãJUNG (Java Universal Network/Graph Framework) ãšãã£ãã©ã€ãã©ãªãå©çšå¯èœã§ãã
ã°ã©ãã¢ããªãã£ã¯ã¹ãå§ããã«ã¯
ã°ã©ãã¢ããªãã£ã¯ã¹ãåããŠåŠã¶æ¹ã®ããã«ãå§ããããã®ã¹ããããããã€ã玹ä»ããŸãïŒ
- åºç€ãåŠã¶ïŒã°ã©ãçè«ããããã¯ãŒã¯åæãã°ã©ãããŒã¿ããŒã¹ã®åºæ¬æŠå¿µãçè§£ããŸãã
- ã°ã©ãããŒã¿ããŒã¹ãéžã¶ïŒããŒãºãšäºç®ã«åã£ãã°ã©ãããŒã¿ããŒã¹ãéžæããŸããå€ãã®ãŠãŒã¶ãŒã«ãšã£ãŠãNeo4jãè¯ãåºçºç¹ã§ãã
- ã°ã©ãã¢ããªãã£ã¯ã¹ããŒã«ãæ¢ãïŒããŸããŸãªã°ã©ãã¢ããªãã£ã¯ã¹ããŒã«ããã©ãããã©ãŒã ã詊ããèªåã®ã¯ãŒã¯ãããŒã«æé©ãªãã®ãèŠã€ããŸãã
- ç°¡åãªãããžã§ã¯ãããå§ããïŒå°èŠæš¡ã§æç¢ºã«å®çŸ©ãããåé¡ã«ã°ã©ãã¢ããªãã£ã¯ã¹ãé©çšããå®è·µçãªçµéšãç©ã¿ãŸãã
- ã³ãã¥ããã£ã«åå ããïŒä»ã®ã°ã©ãã¢ããªãã£ã¯ã¹ã®å®è·µè ãç ç©¶è ãšã€ãªããã圌ãã®çµéšããåŠã³ãèªåã®çµéšãå ±æããŸããã«ã³ãã¡ã¬ã³ã¹ã«åå ãããããªã³ã©ã€ã³ãã©ãŒã©ã ã«åå ãããããªãŒãã³ãœãŒã¹ãããžã§ã¯ãã«è²¢ç®ãããããŸãããã
ã°ã©ãã¢ããªãã£ã¯ã¹ã®èª²é¡ãšä»åŸã®åå
ã°ã©ãã¢ããªãã£ã¯ã¹ã¯å€§ããªå¯èœæ§ãç§ããŠããŸãããããã€ãã®èª²é¡ãæ±ããŠããŸãïŒ
- ã¹ã±ãŒã©ããªãã£ïŒéåžžã«å€§èŠæš¡ãªã°ã©ãã®åæã¯ãèšç®ã³ã¹ããé«ããå°éã®ããŒããŠã§ã¢ãšãœãããŠã§ã¢ãå¿ èŠãšããããšããããŸãã
- ããŒã¿çµ±åïŒè€æ°ã®ãœãŒã¹ããã®ããŒã¿ãäžè²«æ§ã®ããã°ã©ãæ§é ã«çµ±åããããšã¯è€éã«ãªãå¯èœæ§ããããŸãã
- ã¢ã«ãŽãªãºã ã®éžæïŒç¹å®ã®åé¡ã«å¯ŸããŠé©åãªã°ã©ãã¢ããªãã£ã¯ã¹ã¢ã«ãŽãªãºã ãéžæããããšã¯é£ããå ŽåããããŸãã
- çµæã®è§£éïŒã°ã©ãã¢ããªãã£ã¯ã¹ã®çµæãè§£éãããããå®è¡å¯èœãªã€ã³ãµã€ãã«å€æããã«ã¯å°éç¥èãå¿ èŠã§ãã
ã°ã©ãã¢ããªãã£ã¯ã¹ã®ä»åŸã®ååã«ã¯ä»¥äžãå«ãŸããŸãïŒ
- ã°ã©ãæ©æ¢°åŠç¿ïŒã°ã©ãã¢ããªãã£ã¯ã¹ã𿩿¢°åŠç¿ãçµã¿åããããã匷åãªäºæž¬ã¢ãã«ãéçºããŸãã
- ãªã¢ã«ã¿ã€ã ã°ã©ãã¢ããªãã£ã¯ã¹ïŒã°ã©ãããŒã¿ããªã¢ã«ã¿ã€ã ã§åæããå³æã®æææ±ºå®ããµããŒãããŸãã
- 説æå¯èœãªã°ã©ãAIïŒäºæž¬ãæšå¥šã«å¯Ÿãã説æãæäŸããã°ã©ãã¢ããªãã£ã¯ã¹æè¡ãéçºããŸãã
- ãã¬ããžã°ã©ãã®èªååïŒãã¬ããžã°ã©ãã®äœæãšç¶æãèªååããŸãã
çµè«
ã°ã©ãã¢ããªãã£ã¯ã¹ã¯ãæ¥ç¶ãããããŒã¿ã®äžã«é ããããã¿ãŒã³ãé¢ä¿æ§ãã€ã³ãµã€ããæããã«ããããã®åŒ·åãªããŒã«ã§ããã°ã©ãããŒã¿ããŒã¹ãã¢ã«ãŽãªãºã ããã©ãããã©ãŒã ãæŽ»çšããããšã§ãçµç¹ã¯ç«¶äºäžã®åªäœæ§ãç²åŸããè€éãªåé¡ã解決ããå¹ åºãæ¥çã§æ å ±ã«åºã¥ããæææ±ºå®ãè¡ãããšãã§ããŸããããŒã¿ããŸããŸãçžäºæ¥ç¶ãããã«ã€ããŠãã°ã©ãã¢ããªãã£ã¯ã¹ã®éèŠæ§ã¯å¢ãç¶ããã€ãããŒã·ã§ã³ãšçºèŠã®æ°ããªæ©äŒãæäŸããã§ããããã€ãªããã®åãåãå ¥ããã°ã©ãã¢ããªãã£ã¯ã¹ã§ããŒã¿ã®å¯èœæ§ãè§£ãæŸã¡ãŸãããã
ãã®èšäºã§ã¯ãã°ã©ãã¢ããªãã£ã¯ã¹ã®å æ¬çãªæŠèŠãæäŸããŸããããã®åéãé²åããã«ã€ããŠããã®å¯èœæ§ãæå€§éã«åŒãåºãããã«ã¯ãç¶ç¶çãªåŠç¿ãšå®éšãäžå¯æ¬ ã§ããäžæ žãšãªãæŠå¿µãçè§£ããããŸããŸãªæè¡ãæ¢æ±ããææ°ã®ãã¬ã³ããææ¡ããããšã§ãã°ã©ãã¢ããªãã£ã¯ã¹ã®åãæŽ»çšããŠè²Žéãªã€ã³ãµã€ããç²åŸããçµç¹ã«ãšã£ãŠææçŸ©ãªææãæšé²ããããšãã§ããŸãã