D. Kazakov and S. Manandhar. A hybrid approach to word segmentation. In D. Page, editor, Proceedings of the 8th International Conference on Inductive Logic Programming, volume 1446, pages 125--134. Springer-Verlag, 1998.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....the isolated script objects. Essential preprocessing steps are the reconstruction and separation of text lines which have been described in [1] and [2] If information about the characters is available, the analysis of character se quences leads to reliable predictions about word boundaries [3]. However, we don t have this information. In order to find word gaps, several distance measuring methods have been investigated, such as run length Euclidean heuristic distance [10] convex hull [7] etc. These methods analyse relations between ad jacent connected components and find gap metrics ....

D. Kazakov and S. Manandhar. A hybrid approach to word segmentation. In D. Page, editor, Proceedings of the 3th International Conference on Inductive Logic Programming, volume 1446, pages 125 134. Springer-Verlag, 1998.

....through to a default rule if none apply. The ILP system learns a very concise and comprehensible de nition for the past tense transformation using this approach. Similar ILP methods have also been applied to learning morphology in other European languages (Manandhar, Dzeroski, Erjavec, 1998; Kazakov Manandhar, 1998). 20.6.2 Part of Speech Tagging Tagging each word with its appropriate part of speech (POS) based on context is a useful rst step in syntactic analysis (see chapter 11) In addition to statistical methods that have been successfully applied to this task, decision tree induction (Marquez, Padro, ....

Kazakov, D., & Manandhar, S. (1998). A hybrid approach to word segmentation. In Proceedings of the 9th International Workshop on Inductive Logic Programming (ILP-99), pp. 125-134. Springer.

.... grammars (PCFGs) There has been some recent research on logic based language learning (Mooney Califf, 1995; Cohen, 1996; Freitag, 1998) in particular a recent body of European inductive logic programming (ILP) research on language (Cussens, 1997; Manandhar, Dzeroski, Erjavec, 1998; Kazakov Manandhar, 1998; Eineborg Lindberg, 1998; Lindberg Eineborg, 1998; Cussens, Dzeroski, Erjavec, 1999; Lindberg Eineborg, 1999) However, most of this research has focused on relatively low level tasks such as morphological analysis and part of speech tagging and has not conclusively demonstrated ....

Kazakov, D., & Manandhar, S. (1998). A hybrid approach to word segmentation. In Inductive Logic Programming: Proceedings of the 8th International Workshop (ILP-98). Springer.

....is interesting that, due to the use of the database, lists have been rejected in favour of indexing sentences and words. On the plus side this allows the use of the database, on the minus side the size of the data is increased. 2. 4 Morphology ILP approaches to learning morphology can be found in [6, 12, 9]. In all of these, words are represented as lists of letters and a single list processing predicate is used to perform morphological analysis. In [6] inflectional paradigms of Slovene nouns are learned. From examples such as: nxmsg( t,e,s,l,a] t,e,s,l,e] rules such as: nxsmg(A,B) ....

....were learned for a large number of forms, not just masculine genitive. The work in [12] has a similar representation design except the single predicate mate 6 is used where mate(W1,W2,P1,P2,S1,S2) is true if P1 and S1 is a prefix and suffix of W1 and similarly for P2, 21 and W2. Similarly in [9], the single background predicate append 3 is used to learn segmentation rules such as seg(A,B) append( e,b,l,o,u,i,s,s] B,A) append(C, o,n,s] A) from examples such as: seg( e,b,l,o,u,i,s,s,o,n,s] o,n,s] The representational features here are the use of lists and list processing ....

[Article contains additional citation context not shown here]

Dimitar Kazakov and Suresh Manandhar. A hybrid approach to word segmentation. In Inductive Logic Programming: Proceedings of the 8th International Workshop (ILP-98). Springer, 1998.

No context found.

Dimitar Kazakov and Suresh Manandhar. A hybrid approach to word segmentation. In David Page, editor, The Eighth International Conference ILP-98, pages 125--134, Madison, Wisconsin, USA, 1998. Springer-Verlag.

....and Foidl on the task of learning rules for the analysis of English nouns. Training set size 100 200 300 400 500 600 700 800 900 1063 Clog [s] 21 61 112 204 277 477 646 941 1198 1771 Foidl [s] 254 1290 2191 5190 9521 18353 33915 In our earlier work on the segmentation of French verbs (Kazakov and Manandhar, 1998), the running times of Clog on a data set of 200 and 500 training examples were 96 and 811 seconds respectively. For Foidl, they were 900 and 14428 seconds respectively on the same data sets. Our experience with Clog indicates that it is ideally suited for applications where the user has a clear ....

Kazakov, D. and S. Manandhar: 1998, `A Hybrid Approach to Word Segmentation'. In: D. Page (ed.): Proc. of the Eighth International Conference on Inductive Logic Programming. Madison, Wisconsin, pp. 125--134, Springer-Verlag.

No context found.

D. Kazakov and S. Manandhar. A hybrid approach to word segmentation. In D. Page, editor, Proceedings of the 8th International Conference on Inductive Logic Programming, volume 1446, pages 125--134. Springer-Verlag, 1998.

No context found.

Dimitar Kazakov and Suresh Manandhar. 1998. A hybrid approach to word segmentation. In David Page, editor, Proceedings of the ILP-98. Springer. Lectures Notes in Computer Science, vol. 1446.

No context found.

Dimitar Kazakov and Suresh Manandhar. A hybrid approach to word segmentation. In David Page, editor, Proceedings of the ILP-98. Springer, 1998. Lectures Notes in Computer Science, vol. 1446.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC