Raka'a
1. Gabatarwa zuwa Haƙiƙa na Artificial a Chemistry: Matsayi, Iyakoki, Tabbatarwa da Da'a 2. Wakilin Kwayoyin Halitta da Tsarin Sinadarai: Tabbatarwa tare da SMILES, InChi da RDKit 3. Hasashen amsawa da injina: Samfur, yanayi da Hasashen Hasashen Mummuna 4. Taimakon Fassarar Bakan: NMR, IR da Mass Spectrometry 5. Bita na wallafe-wallafe da Tsare-tsaren Haɗa: Tushen, Tsari da Taswirar Hanya 6. Binciken Bayanai na Chemical da Python: pandas, Stoichiometry da Calibration 7. Hasashen Dukiyar Kwayoyin Halitta da QSAR: Ƙaddamarwa, pKa da Iyakan Model 8. Takardun dakin gwaje-gwaje: Littafin Bayanin Gwaji, ELN da Maimaituwa 9. Ƙimar Tsaro da Haɗari: GHS, SDS da Iyaka na Hankali na Artificial 10. Tabbatarwa, Hallucination, da Mutuncin Kimiyya 11. Dan Adam a Gwaji: Da'a, Keɓantawa, Nauyi da Ƙarshe-zuwa Ƙarshen Aiki
Naúrar 5 / 11

Bita na wallafe-wallafe da Tsare-tsaren Haɗa: Tushen, Tsari da Taswirar Hanya

Riba:

  • Ƙarfin sanya hankali na wucin gadi azaman kayan aiki mai haɗari a cikin samar da ƙaƙƙarfan ƙaƙƙarfan ƙididdiga a cikin ra'ayoyi, abstracts da dabarun bincike.
  • Ability don kawar da haɗarin tushen karya ta hanyar tabbatar da kowane magana da DOI a cikin bayanan
  • Ikon kafa ingantaccen aikin aiki wanda mutane ke samun ainihin labarin kuma su taƙaita rubutun zuwa hankali na wucin gadi.

Bita na wallafe-wallafe da Tsare-tsaren Haɗawa: Tushen, Tsari da Taswirar Hanya tare da AI

Kowane aikin sinadarai yana farawa da nazarin wallafe-wallafe: "Shin an yi wannan fili a baya? Ta wace hanya? Tare da wane yawan amfanin ƙasa?" AI babban mai haɓakawa ne a wannan matakin; yana taƙaita batun, ya bayyana mahimman ra'ayoyi, samar da kalmomin bincike, tsara hanyoyin. Amma kuma shine tushen kuskure mafi haɗari a cikin ilmin sunadarai: ƙiyayya. AI na iya ƙirƙirar labarai, marubuta, da DOIs waɗanda suke kama da gaskiya amma babu su kwata-kwata. A cikin wannan rukunin, za mu koyi yin amfani da AI a cikin wallafe-wallafe da tsarawa, amma ba tare da barin horo na ƙididdiga ba.

Me yasa ladabtarwa ke da mahimmanci?

Samfurin harshe yana samar da rubutu; baya tambayar ma'ajin bayanai (sai dai in an bayyana akasin haka). "Journal of Organic Chemistry, 2018, 83, 4521, Zhang et al." Abubuwan da ake amfani da fom shine kawai "saitin kalmomi da lambobi waɗanda suke da alama zai yiwu" don ƙirar. Wannan tushen yana iya ko ba zai zama na gaske ba; samfurin ba ya bambanta tsakanin su biyun. Saboda haka, babu wani ambaton da YZ ya bayar da ake amfani da shi ba tare da an tabbatar da shi ba a cikin ma'ajin bayanai (mai nazarin DOI, PubMed, Crossref, wurin wallafawa).

Hankali: Kawai saboda AI yana ba da DOI baya nufin cewa DOI yana da inganci. A warware DOI ta https://doi.org/; Idan bai kai ga labarin ba, wannan tushen ba ya wanzu, ko da lakabi da marubucin suna da alama na gaske.

Inda AI ke da ƙarfi da rauni a cikin wallafe-wallafe

  • Ƙarfi: Bayyana ra'ayi, zayyana nau'ikan halayen halayen, samar da kalmomin bincike, bayyana manyan matakai na hanya, taƙaita rubutun labarin (lokacin da kuka liƙa).
  • Rauni: Ba da "daga ƙwaƙwalwar ajiya" wanzuwar takamaiman labarin, ainihin shafi / ƙarar / bayanin DOI, daidai ƙimar yawan amfanin ƙasa. Waɗannan su ne wuraren da aka fi buɗe don ƙirƙira.

Wannan bambance-bambancen yana haifar da dabarun da ke gaba: amfani da AI azaman kayan aikin sarrafa rubutu (takaitawa, rarrabawa), ba azaman tushen gaskiya ba. Kuna sami ainihin labarin, sa AI ya karanta kuma ku taƙaita shi.

Mataki-mataki: amintaccen adabi + kwararar tsari

  1. Koyi manufar: Shin AI ta bayyana batun, ajin amsa, sharuɗɗan. Haɗarin ƙirƙira yana da ƙasa a nan.
  2. Ƙirƙirar kalmar bincike: Tambayi AI don kalmomi 8-10 / kalmomi masu ma'ana da yiwuwar mujallolin; Bincika ainihin bayanan bayanai tare da waɗannan.
  3. Samun ainihin labarin: Nemo labarin da kanku daga PubChem/Reaxys/Google Scholar.
  4. Taƙaitawa: Manna rubutun labarin cikin AI kuma yin taƙaitaccen tsari (hanyar, yanayi, yawan amfanin ƙasa, iyakancewa).
  5. Yi tsari: Ƙirƙiri taswirar haɗakarwa don manufa; haɗa kowace gudu zuwa ainihin tushen.
  6. Tabbatarwa: Gyara kowace magana tare da DOI; Kwatanta kowane lamba mai mahimmanci tare da rubutun tushe.

Samfura huɗu masu kwafi

1) Ra'ayi + Ƙirƙirar kalmar bincike (yanki mai aminci):

Maudu'i: Ƙaddamar da haɗin aryl-aryl ta hanyar Suzuki-Miyaura coupling reaction.Task: 1) Bayyana abin da ke faruwa a cikin jimloli 5 (ma'anar kalmomi).2) Ƙirƙirar kalmomi / kalmomi na 10 don binciken wallafe-wallafe. Zan neme su a cikin rumbun adana bayanai.

2) Takaitacciyar labarin (kun bayar da rubutun):

A ƙasa ina liƙa ɓangaren gwaji na labarin.[TEXT] Aiki: Taƙaice tare da waɗannan jigogi: - Dandalin manufa da mahimmancinsa- Hanyar/reagents da aka yi amfani da su- Yanayi (zazzabi, lokaci, ƙarfi) - Haɓaka- Ƙimar da aka bayyana Kada a ƙara kowane lambobi BA a cikin rubutu ba; in ba haka ba rubuta "wanda ba a bayyana ba".

3) Taswirar hanyar haɗin gwiwa (dangane da tushe):

Manufa: 2-phenylbenzimidazole (SMILES: c1ccc (cc1) c1nc2ccccc2 [nH] 1) .Task: Ba da daftarin tsarin haɗin 2-mataki. Ga kowane mataki: reagents, yanayin, hanyar tsarkakewa.Dokar: Ga kowane mataki, ƙara "nau'i na yau da kullum na wannan nau'i na amsawa tare da wace kalma don bincika bayanai DOI FAKE.

4) Tabbataccen lissafi:

Akwai nassoshi 4 a ƙasa. Samar da lissafin tantance gaskiya ga kowane: waɗanne fage (DOI, mujallu, juzu'i, shekara, marubuci) zan iya tantancewa kuma a ina? Kawai gaya mani mataki-mataki yadda zan tabbatar.

Rauni mai ƙarfi / Ƙarfi mai ƙarfi

Rauni:

Ka ba ni tushe guda 5 akan haɗin wannan fili.

Mai yiwuwa AI ta samar da haƙiƙanin halayen amma ba a tantance su ba, wasu daga cikinsu ƙirƙira ne.

Mai ƙarfi:

Maudu'i: Ƙirƙirar fili Zan nemo ainihin maɓuɓɓuka daga rumbun adana bayanai; Kuna ƙarfafa dabarun bincike na.

Bambanci: Mun jagoranci AI zuwa aikin da yake da ƙarfi (dabarun bincike, ra'ayi) kuma mun kiyaye shi daga aikin da ke da sauƙin ƙirƙira (ƙirar ƙira).

Nau'in tushe da dogaro

Tushen/kayan aiki

Me ke bayarwa

dogara

Matsayi

AI (samfurin harshe) daga ƙwaƙwalwar ajiya

Ra'ayi, shaci, ra'ayin nema

Babban ra'ayi, ƙananan kan tunani

ra'ayi / taƙaitawa

AI + rubutu da aka liƙa

Takaitaccen rubutun da aka bayar

Babban (dangane da rubutu)

taƙaitawa

DOI parser (doi.org)

Sahihancin sifa

babba

Tabbatarwa

PubChem / Reaxys / SciFinder

fili, dauki, tushen

babba

bincike na farko

Google Scholar / Crossref

Neman labarin, ambato

babba

tabbatarwa

kananan lokuta

Shari'a ta 1 - Tushen ƙirƙira guda uku. Wani mai bincike ya tambayi AI don tushen 6 akan wani batu. Lokacin da ya warware DOIs daya bayan daya, 3 daga cikinsu ba su da inganci; Kanun labarai sun kasance masu gamsarwa, amma labaran ba su kasance ba. Tabbacin ya ɗauki mintuna 25 amma ya hana ƙididdiga 3 na karya shiga labarin. Darasi: Ana warware kowace magana ɗaya bayan ɗaya.

Case 2 - Ƙarya yawan amfanin ƙasa. "Yin aiki shine 92%," in ji AI lokacin da ke bayyana hanya. Lokacin da mai amfani ya samo kuma ya kalli ainihin labarin, yawan amfanin ƙasa shine 64%; AI ta sanya lambar daga ƙwaƙwalwar ajiya. Darasi: kwatanta mahimman lambobi tare da rubutun tushe, kar a amince da taƙaitawar AI.

Case 3 - Darajar a cikin ra'ayi, haɗari a cikin tunani. Wani dalibi yana da AI ya bayyana ajin dauki; Bayanin ya kasance daidai da koyarwa. Sannan "manyan labarai 3?" sai ya ce; Biyu daga cikinsu sun kasance ƙage ne. AI duka biyun suna da amfani sosai kuma suna da haɗari a cikin wannan zaman. Darasi: abin hawa ɗaya yana nuna aminci daban-daban a cikin ayyuka daban-daban; Amince gwargwadon aikinku.

Kuskuren gama gari

  • Amfani da nassoshi ba tare da tabbatarwa ba. Kuskuren da ya fi kowa kuma mafi tsanani; na iya haifar da ja da baya labarin.
  • Samun yawan amfanin ƙasa/lambobin yanayi daga taƙaitawa. Takaitaccen bayanin AI wani lokaci yana nuna "ƙimar mai yiwuwa" maimakon tushen; kwatanta da asali.
  • Kuskure AI don injin bincike. Sai dai in an faɗi akasin haka, ƙirar yare ba ta neman bayanan kai tsaye; yana samarwa daga ƙwaƙwalwar ajiya.
  • Bincike a cikin harshe ɗaya. Kawai barin mahimman kalmomi daga Turkanci/Ingilishi ya ɓace albarkatun; AI yana taimakawa a cikin tsararraki masu ma'ana.
  • Ra'ayi mai ruɗani tare da tushe. Gaskiyar cewa AI ya bayyana madaidaicin ra'ayi baya nuna cewa bayanin da yake bayarwa daidai ne.
Tukwici: Mafi ƙarfin aikin aiki shine "ɗan adam na farko ya samo, sannan AI ta taƙaita." Kuna kawo ainihin labarin kuma ku sa AI ta karanta rubutun. Don haka kuna amfani da ƙarfin AI (takaice) kuma ku kashe rauninsa (daidaituwar tushen).

A takaice

  • AI yana da haɗari wajen samar da ƙaƙƙarfan ƙididdiga masu ƙarfi a cikin ra'ayoyi, taƙaitawa, da dabarun bincike a cikin adabi.
  • Ƙirar ƙarya (labarin ƙirƙira / DOI) shine mafi girman haɗarin AI a cikin sunadarai; An tabbatar da kowace magana.
  • Mafi aminci kwarara: ɗan adam ya sami ainihin labarin, AI ta taƙaita rubutun.
  • Ɗauki lambobi masu mahimmanci, kamar yawan amfanin ƙasa da yanayi, daga rubutun tushe, ba ƙididdiga ba.
  • Haɗa kowane mataki a cikin shirin haɗawa zuwa tushen asali.

Aikin aikace-aikace

Zaɓi mahallin manufa. Da farko, tambayi AI don sharuɗɗan bincike da taƙaitaccen ra'ayi kawai (ba tare da neman takamaiman sifa ba). Nemo aƙalla labarai biyu a cikin ainihin ma'aunin bayanai tare da waɗannan sharuɗɗan. Takaita rubutun labarin zuwa AI kuma kwatanta yawan amfanin ƙasa/lambobi zuwa na asali. Na dabam, da gangan ka tambayi AI don "maɓuɓɓuka 5" kuma ka gano DOIs yana ba da ɗaya bayan ɗaya: nawa ne na gaske? Rubuta "rahoton tabbatar da magana" shafi daya.

jerin abubuwan dubawa

  • [ ] Na sanya AI azaman ra'ayi / kayan aiki na taƙaitawa, ba injin bincike ba.
  • [ ] Lokacin da nake son yin magana, na yanke kowane DOI daga doi.org.
  • [ ] Ina kwatanta lambobin amfanin ƙasa/sharadi zuwa tushen asali.
  • [ ] Na sami ainihin labarin kuma ina sa AI ta taƙaita rubutun.
  • [ ] Na dangana kowane mataki a cikin shirin hadawa zuwa tushe.
  • [ ] Na shirya aƙalla rahoton tabbatarwa guda ɗaya.
A baya Na gaba