Our corpus level 2 consists of the references in the notes of articles. It also follows TEI guidelines as the corpus level 1, but needs another operation concerning extraction of reference part. That is, since the note contents are the mix of bibliographic information and other information, we first manually extract valid parts for references.

The problem here is that we need exactly the same operation in an automatic way on new references to be estimated. By enumerating in detail the difficulties for the extraction, we would find the right way to approach this problem.

1. Not all the numbered notes have bibliographic information while we need to take only the notes having this information. Actually, the valid notes for references are selected during the manual annotation of corpus level 2. But we should develop a way to automate the selection process for the estimation of new references in notes.

-Example———————————————————————————————————————

26. La nature euro-centrée du projet était encore plus apparente dans la version originale du texte, qui, comme nous l’avons déjà mentionné, était appelé « Europe élargie ».

27. « Une Europe sûre dans un monde meilleur. Stratégie européenne de sécurité », Bruxelles, 12 décembre 2003. Pour un commentaire critique, voir Toje A., « The 2003 European security strategy: A critical appraisal », European Foreign Affairs Review, vol. 10, n°1, 2005, pp. 117-134.

——————————————————————————————————————————–

2. A lot of notes have some additional phrases besides bibliography parts. We also need to automatically extract the appropriate parts containing only bibliographic information.

-Example———————————————————————————————————————

31. Voir par exemple la communication de la Commission relative au « Renforcement de la politique européenne de voisinage », COM (2006) 726 final.

32. Pour une lecture des approches sociologiques de l’étude de l’Europe et de l’UE, voir Christiansen T., Joergensen K., Wiener A. (eds), « Constructivism in European studies », European Journal of Public Policy, n°6, 1999.

——————————————————————————————————————————–

3. There are many notes that consist of a term indicating a previously cited source, and some essential information for the reference.

-Example———————————————————————————————————————

12. Del Sarto et Schumacher qualifient cette approche de « bilatéralisme différencié », voir Del Sarto R.A., Schumacher T., op. cit., p. 5.

13. COM (2003) 104 final, op. cit., pp. 11-12.

(…)

33. Swidler A., « Culture in action: symbols and strategies », American Sociological Review, n°51, 1986.

34. Ibid.

——————————————————————————————————————————–

4. Sometimes, an article has both bibliography and note parts. The former consists of a set of well-formed references and the latter includes less formulaic references as explained above. Moreover, we can find several references in note part indicating other references in bibliography part. Xml identifiers are added to make links between the two parts during the construction of corpus 2. In this situation, we need to decide how to handle these links for learning and estimation. For example, if we want to ignore the links between two parts, we just take data in note part for learning.

-Example———————————————————————————————————————

Bibliography

Aylwin, José, « Los ombudsman y los derechos de los pueblos indígenas en América Latina ». In : Míkel Berraondo (coord.), 2006, p 339-357.

(…)

Quijada, Mónica, « Indígenas, violencia, tierra y ciudadanía ». In : M. Quijada, C. Bernand et A. Schneider (éds.), Homogeneidad y nación. Con un estudio de caso : Argentina, siglos XIX y XX, Madrid, CSIC, 2000, p. 57-92.

Quijada, Mónica, 2006, « L’Argentine à travers le miroir. Construction et dé-construction d’une élaboration identitaire collective ». In : J.-P. Castelain, S. Gruzinski et C. Salazar Soler (éds.), De l’ethnographie à l’histoire. Paris – Madrid – Buenos Aires. Les mondes de Carmen Bernand, Paris, L’Harmattan, 2006, p. 329-348.

Notes

(…)

[4] Aylwin 2006, p. 339-342.

[5] Quijada, 2006, p. 338. M. Quijada est par ailleurs l’auteur d’une remarquable analyse historique sur l’intégration des amérindiens dans la société argentine (Quijada, 2000).

——————————————————————————————————————————–

---

---